The Blackbird Brief

Portfolio Watch — the most consequential thing in yesterday's pancreatic-cancer approval was not the survival number, it was the absence of a companion diagnostic: FDA cleared Revolution Medicines' RASONQUE (daraxonrasib), an oral RAS(ON) tri-complex molecular glue, in previously treated metastatic pancreatic adenocarcinoma on OS of 13.2 vs 6.7 months (HR 0.40) — with NO required RAS mutation test, ~5 weeks from submission and 6.5 months early via the Commissioner's National Priority Voucher, at $39,800 a month. So the new second-line default has zero patient-selection steps while all three Blackbird mutant-KRAS positions are mutation-selective (Clasp needs KRAS G12V AND HLA-A*03:01; Adventris a defined mutant peptide; aSKY its own frame) — the KRAS map needs an eligible-fraction row, and Clasp's resistance positioning just became an addressable market with a start date 7.2 months out. Plus Spyre's anti-TL1A missing its own bar in rheumatoid arthritis pushes IBD toward combinations, which is the case for rewriting the Slusher gut-restricted GCPII program as a combination-partner story; the voucher's two proof points priced at opposite extremes ($39,800/month vs Regeneron giving Otarmeni away free in Aletira's exact indication); Aurora Therapeutics killed by a competitor's press release seven months after a $16M Doudna/Urnov launch; and SK paying Biohaven $400M cash up front for a de-risked Kv7 channel at Phase 2/3 while Nxera's novel GPR52 asset still has no buyer

Thu, 27 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Thursday, August 27, 2026. LEAD (2026-08-26): FDA approved RASONQUE (daraxonrasib, Revolution Medicines) for adults with metastatic pancreatic adenocarcinoma after at least one prior systemic therapy or who are not candidates for multiagent therapy — an oral RAS(ON) multi-selective non-covalent tri-complex inhibitor that glues RAS to cyclophilin A and blocks effector engagement, active across wild-type RAS and G12D/G12V/G12R/G12C. RASolute 302 (n=500 vs SOC chemo): OS 13.2 vs 6.7 months, HR 0.40 (0.30-0.53), p<0.0001; PFS 7.2 vs 3.6 months, HR 0.49; pain time-to-deterioration 9.2 vs 3.8 months. Safety is rough (rash 86%, diarrhea 63%, stomatitis 57%, fatal GI perforation and fatal ILD) yet only 2.9% discontinued permanently. NO COMPANION DIAGNOSTIC — approved with or without an identified RAS mutation. $39,800 per 30-day supply; submitted 2026-07-22, approved 2026-08-26, 6.5 months ahead of the goal date via the FDA Commissioner's National Priority Voucher; RBC peak sales ~$11.5B. THREE BLACKBIRD READS: (1) the biomarker-free label, not the hazard ratio, is the story — the new second-line default has zero patient-selection steps while Clasp's CLSP-5282 requires KRAS G12V AND HLA-A*03:01 (two gates), Adventris needs a defined mutant-KRAS peptide and aSKY its own frame, so the KRAS competitive map built 08-14/08-15 needs an ELIGIBLE-FRACTION row before the next partnering conversation. (2) The resistance market now has a start date: Clasp's AACR 2026 positioning (kills harder against KRAS-inhibitor-resistant tumors, synergistic in combination) converts from slide to market, and at 7.2 months of PFS the first US post-RAS-inhibitor PDAC cohort arrives around spring — Eddie should ask whether SENTINEL-101 carries a post-daraxonrasib stratum. (3) Of the five-column KRAS board, the small-molecule column just took the largest indication with chemistry rather than immunology, so the bar for a novel immunologic modality is now what it adds that the pill cannot do. ITEM 2 (2026-08-25/26): Spyre's SPY072 anti-TL1A beat placebo on DAS28-CRP at the low dose in the 143-patient SKYWAY rheumatoid arthritis sub-study but missed at the high dose and fell short of Spyre's own bar for monotherapy; RA dropped, stock -13% and over $1B of market value gone, PsA and axSpA due Q4, UC through 2027-2028, sell-side calling the IBD combination thesis intact. The class carrying $10B+ of deal value just signalled it may not carry a monotherapy label outside the gut, which makes an oral, gut-restricted, non-immunosuppressive partner the most valuable thing in IBD — a description of Barbara Slusher's GCPII inhibitor, whose positioning document should be rewritten as a combination-partner story. ITEM 3: the Commissioner's National Priority Voucher now has two proof points across two therapeutic areas and two modalities — Regeneron's Otarmeni (61 days from BLA, April 23) and RASONQUE (about five weeks) — but they priced at opposite extremes, $39,800 a month against Otarmeni provided FREE in the US. For Aletira that is a demonstrated 61-day precedent in its own indication and unclaimed NPV, alongside a list-price-of-zero anchor that the Series A narrative has to answer before the round opens (GJB2 is a far larger population than OTOF; that is the answer). ITEM 4: Aurora Therapeutics, the first bespoke gene-editing company, co-founded by Jennifer Doudna and Fyodor Urnov with a $16M Menlo seed in January, killed its lead PKU program and cut staff seven months in after Beam announced BEAM-304 in the same disease six weeks post-launch — killed by a competitor's press release, not by its own data, on three times the current mean seed round. Every venture partner should name the specific well-capitalized group most likely to preempt each active program and date when they would have to move. ITEM 5: SK Biopharmaceuticals paid Biohaven $400M cash up front ($350M at closing, $50M at one year), up to $395M in milestones and mid-teens to low-twenties US royalties for the Kv7 platform and opakalim (Kv7.2/7.3 activator, Phase 2/3 focal epilepsy), plus up to $245M of obligations to originator Knopp Biosciences. CNS money is real, but it went to a de-risked channel at Phase 2/3 rather than a novel GPCR at Phase-2-ready, while Nxera's NXE'149 still has no named buyer nine months after Boehringer opted out — so for the Lieber GPR52 program the development-candidate declaration is not the value inflection; first-in-human receptor occupancy is. ITEM 6: Faro's $37.3M Series B co-led by Merck Global Health Innovation Fund and S32 confirms strategic money underwrites trial cycle time, but it bought the protocol and structured-data layer rather than enrollment — the lane 1104health occupies is still open. Nulls: zero GPR52 items, no Nxera buyer, nothing in hearing loss since Skylark dosed on 08-11, nothing from CAMP4 since the 08-13 quarter, silence from Clasp, Adventris, aSKY and Sensorion, no JHU/UMB/Lieber/UM Ventures item in the window, and zero Blackbird mentions.

Portfolio Watch — Johns Hopkins invents at industrial scale and capitalizes at artisanal scale, and the gap is exactly where Blackbird lives: JHTV's FY2026 report shows nearly $100M revenue, $65M of it licensing concentrated in just two products (Pylarify, approved 2021, and vorasidenib, approved 2024), 462 invention reports converting to 17 startups (3.7%), and 142 active startups raising $151M in a year — about $1.06M per company, less than two Massachusetts Series A rounds, because MassBio's snapshot the same day put the mean Series A at $79.6M while the average seed collapsed to $4.65M from $7.65M a year earlier. The capital barbell in one line: 25 biotech IPOs YTD and the M&A freeze declared over, growth rounds enormous, seed starving nearly 40% in a year total venture grew 25%. What it means for Aletira's Series A (Leerink's two yellow flags are the two ways an incubated company misprices itself), for the Blackbird model (the nondilutive grant IS the seed the market stopped writing), and for recruiting (Boston shed 3,600 life-sciences jobs, its first decline in two decades, while its capital rose 25%). Plus JHTV's nine translational-fund awards as a free pre-competitive longlist — Schneck's in vivo CAR-T via targeted polymeric mRNA nanoparticles, Won Jin Ho's immune-resistant-tumor immunotherapy, and Yun Guan's MRGPRX1 allosteric modulators for non-addictive analgesia, landing in the same fortnight roughly $395M went into pain assets on two in-licensed molecules neither buyer discovered

Wed, 26 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Wednesday, August 26, 2026. LEAD: Johns Hopkins Technology Ventures published its FY2026 annual report (posted 2026-08-17; Johns Hopkins Hub coverage 2026-08-25) covering the year ended 2026-06-30 — nearly $100M total revenue, $65M licensing revenue, 462 invention reports, 111 new U.S. patents against ~3,500 active patents, 120 licensing agreements, 349 products on market, 98 active research-based corporate partnerships, 17 new startups formed, 142 active startups, and $151M in venture capital raised by JHU-affiliated startups with $116M (77%) staying in Baltimore. The majority of licensing revenue comes from two products: Pylarify (prostate PET imaging, from Martin Pomper's chemistry, FDA-approved 2021) and vorasidenib/Voranigo (low-grade glioma, FDA-approved 2024). Three original reads: 17 of 462 is a 3.7% disclosure-to-startup conversion and the constriction sits exactly at turning a disclosure into an investable package with an operator attached, which is the Blackbird thesis stated as a ratio; $151M across 142 companies is about $1.06M per company per year, less than two Massachusetts Series A rounds; and revenue concentrated in two products approved in 2021 and 2024 on science licensed a decade-plus earlier means JHTV's live incentive is sourcing the next Pylarify, not extracting royalty on a preclinical asset. ITEM 2, the capital barbell (both events 2026-08-25): MassBio's 2026 Industry Snapshot has Massachusetts biopharma raising $3.45B in H1 venture, up 25%, while the average seed round fell to $4.65M from $7.65M a year earlier and the mean Series A ballooned to $79.6M; Massachusetts also shed ~3,600 life-sciences jobs in 2025, its first annual decline in more than two decades, with R and D employment down 3.9%. Leerink's Jack Bannister counts 25 biotech IPOs YTD, five in August, the busiest since 2021, and names two yellow flags — preclinical companies rushing to go public and teams overvaluing themselves. Consequences: Aletira's pre-IND package is now the financing inflection, not just regulatory hygiene; a $4.65M market seed no longer buys a package that clears an $80M Series A, so Blackbird Laboratories' nondilutive grant is the missing bridge; and decoupled Boston capital and payroll is a time-limited hiring window for BioHub benches at City Garage. ITEM 3: JHTV's nine translational-funding awards (2026-08-16) across the Cohen, Thalheimer and Zizic funds are a pre-publication longlist of what Hopkins itself thinks is closest to translation — three sit on Blackbird theses, including Jonathan Schneck's in vivo CAR-T via targeted polymeric mRNA nanoparticles, Won Jin Ho's immunotherapy against immune-resistant tumors, and Yun Guan's MRGPRX1 allosteric modulators for non-addictive analgesia, which lands in the same fortnight roughly $395M went into pain assets via two Nasdaq reverse mergers on molecules in-licensed from Italy and China.

Portfolio Watch — the FDA flexibility I sold you yesterday partly unwound inside eighteen hours: REGENXBIO's Hunter syndrome gene therapy is back on clinical hold after spine MRI found a small nodule or cystic mass in 5 of 48 participants dosed intracisternally or intraventricularly three to six years ago, killing the Q3 resubmission that FDA agreed in June needed no additional studies and no untreated control arm. The reason anyone was imaging spines: in January an asymptomatic five-year-old developed an intraventricular tumor four years after the sibling Hurler program, with an AAV integration event tied to overexpression of the proto-oncogene PLAG1, and FDA held both programs on shared-platform risk. Consequences for Aletira: the promoter-to-3'UTR-to-splicing ladder governs where the transgene is read, not where the vector lands, so expression control is not a genotoxicity answer either; the failure mode has a three-to-six-year latency no hearing program will have imaged at decision time; and a written Type A agreement is not durable against new safety data, which moves the long-term safety dataset ahead of the endpoint on the pre-IND agenda. Plus Capricor's PDUFA extended to November 22 on a major amendment carrying 24-month HOPE-3 extension data and a refined upper-limb-function indication — rigorous on evidence, flexible on path, with Kaos Capital's single-asset concentration critique attached — and the FDA commissioner nominee I missed on the 19th, Heidi Overton, who trained at Johns Hopkins

Tue, 25 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Tuesday, August 25, 2026. LEAD (event 2026-08-24): FDA placed a clinical hold on REGENXBIO's RGX-121 (AAV9, MPS II / Hunter syndrome) after asymptomatic spine MRI findings — a small nodule or small cystic mass — in 5 of 48 CAMPSIITE participants dosed intracisternally or intraventricularly approximately three to six years earlier. Findings deemed nonserious and likely benign; no brain masses; no clinical or pathological evidence of nature or causation; spine MRI is not routine in MPS II so background prevalence is unknown. REGENXBIO will not resubmit the BLA in the near term, abandoning the guided Q3 2026 refiling that FDA agreed in June required no additional studies and no untreated control arm; stock fell 22% premarket. CEO Curran Simpson: the findings "are unique and limited to our Hunter Syndrome program." THE UPSTREAM EVENT: on 2026-01-28 the sibling RGX-111 program (MPS I / Hurler) was held after an asymptomatic five-year-old developed an intraventricular CNS tumor four years after intraventricular dosing, with an AAV vector integration event linked to overexpression of the proto-oncogene PLAG1; FDA held both programs on shared-platform-risk grounds, and that surveillance is what found these five. BLACKBIRD READ, correcting this brief's own framing twice: (1) the 08-20 ladder (constitutive promoter → tissue-specific promoter → native regulatory element / Ultragenyx promoter and Skylark 3'UTR → SELEXON splicing) governs where the transgene is read, not where the vector integrates — alongside "expression control is not an immunology answer," add "it is not a genotoxicity answer either"; the surviving claim is the narrow one, that SELEXON solves off-target expression. (2) Latency is 4 years to the tumor and 3–6 years to the spinal findings, versus Skylark SONIX primary completion 2027-12 in ten children — so Geoff Lynn and Hugh Wells should put long-term imaging and neoplasm surveillance into the pre-IND package proactively and price long follow-up into the raise. (3) Yesterday's "get it in writing" is amended: REGENXBIO had written Type A minutes and they are worth nothing this morning because the safety database moved, so negotiate the long-term safety dataset ahead of the endpoint. (4) Eddie: January's dual hold was explicitly shared-platform risk — for any AAV company, ask what else runs on this capsid and whose finding can stop ours. ITEM 2: Capricor's deramiocel PDUFA extended 2026-08-22 to 2026-11-22 after CBER accepted a major amendment with 24-month HOPE-3 open-label extension data and a refined proposed indication focused on upper limb function, the endpoint HOPE-3 actually hit; stock +16%; the STAT rejection scoop did not materialize; Kaos Capital (2026-08-21) demanded board change and a capital-preservation plan, with cash down $80.2M to $237.9M — a single-asset concentration question that applies to Aletira. ITEM 3, coverage gap corrected: on 2026-08-19 Trump nominated Heidi Overton, MD, PhD, for FDA commissioner — general surgery residency at Johns Hopkins, PhD in clinical investigation from the Bloomberg School — though RBC flags no drug-regulatory experience, and the Office of Therapeutic Products vacancy is unchanged. Quiet elsewhere: no RAS catalyst, zero GPR52 items and still no Nxera buyer, nothing from CAMP4 since 08-13, silence from Clasp, Adventris, aSKY, Skylark and Sensorion, and no Blackbird mentions.

Portfolio Watch — the flexibility gene-therapy sponsors are enjoying right now is being granted by an office nobody permanently runs: CBER's Office of Therapeutic Products has had no permanent director since June 2025, its acting director Vijay Kumar left at the end of June, and acting CBER chief Karim Mikhail is running the center and the office at once. That is the constraint on yesterday's rule for the SELEXON pre-IND — align early is worth little if it isn't in writing, because the person you aligned with is temporary. Today's only trade item, a BioSpace analysis, walks the evidence: Replimune approved on the third try four days late on a 24.2% response rate a 10-3 adcomm backed over messy data, uniQure and REGENXBIO both reversed into accelerated-approval filings this quarter, Disc Medicine resubmitting — while Capricor's action date has now passed in silence with two unconfirmed readings live. Plus the second pain-asset reverse merger in five days: Slate takes Fulcrum's Nasdaq shell with $245M for a dual PACAP/VIP antibody in-licensed from China's DartsBio that has not been in a human, which with Ambros puts roughly $395M into pain in a week and none of it into a self-discovered molecule

Mon, 24 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Monday, August 24, 2026. LEAD: the only trade-press item published today is a BioSpace analysis by Tristan Manalac arguing FDA under Kyle Diamantas is restoring reliability after the Makary/Prasad period, with the caution that this is not leniency. Evidence walked: Replimune's TUDRIQEV (vusolimogene oderparepvec-wtpg) plus nivolumab won accelerated approval 2026-08-06 in advanced cutaneous melanoma after progression on an anti-PD-1 regimen — third submission after two rejections, four days past the target date, on a 24.2% objective response rate and 14.1-month median duration of response in 91 efficacy-evaluable patients of 140 in IGNYTE, with a 10-3 advisory committee vote in favor over a dataset the agency itself called messy. uniQure: FDA reversed its demand for another trial and agreed Phase 1/2 data on AMT-130 in Huntington's can support an accelerated approval filing conditioned on a Phase 3 confirmatory, submission guided to Q3. REGENXBIO: the February CRL on the Hunter syndrome gene therapy reversed in June with no additional studies and no untreated control arm required, resubmission also Q3. Disc Medicine: resubmission agreement on bitopertin after a February rejection. Mizuho's Uy Ear — decisions "remain highly data-driven," reconsiderations "highly case-specific," and "the key lesson is to minimize areas open to interpretation." Former CBER staffer Donald Fink — agreement reached about study design "does not ensure regulatory approval." ORIGINAL WORK AND THE ACTIONABLE PART: three of the four reversals ran through one office, CBER's Office of Therapeutic Products, which reviews every gene-therapy IND and BLA and has had NO PERMANENT DIRECTOR SINCE JUNE 2025. Acting director Vijay Kumar stepped down at the end of June 2026; Karim Mikhail, himself acting director of CBER since 2026-05-20, took the office on top of running the center, and the permanent role remains unfilled as of today. Consequence for Hugh Wells and Avi Khanna on the SELEXON pre-IND: verbal alignment with an acting official does not survive that official's replacement, so get the endpoint and the statistical analysis plan into a formal written response, not a good phone call. For the Series A, the reversals are a climate argument — but frame it as "the path exists and is precedented," not "the agency has gotten easier." CAPRICOR, STATED PRECISELY: the 2026-08-22 action date passed with no announcement from company or agency, and two unconfirmed readings are live — STAT reported 2026-08-20 that the therapy is heading for a rejection, while CEO Linda Marban said on the 08-13 call that FDA is receptive to adding open-label extension data, which mechanically implies an extension. Yesterday's lesson on analysis-plan provenance holds either way. SECOND ITEM (event 2026-08-17, not previously covered): Fulcrum Therapeutics and Slate Medicines announced an all-stock reverse merger with an oversubscribed $245M concurrent private placement; Fulcrum holders take ~5%, the company operates as Slate Medicines on Nasdaq as SLTE, funded into 2029. Lead asset SLTE-1009 is a monoclonal antibody binding both PACAP and VIP for migraine prevention, licensed from China-based DartsBio Pharmaceuticals, entering Phase 1 in healthy volunteers in Australia with preliminary data in mid-2027. Paired with Thursday's Ambros/Werewolf deal ($150M PIPE for neridronate, in-licensed from Abiogen), that is roughly $395M into pain assets in five days, both through reverse mergers into Nasdaq shells, neither on a self-discovered molecule. For the gut-restricted GCPII program with Barbara Slusher's team: what the money buys in pain is a stratification package, not a mechanism — Ambros bought enrollment gates, Slate bought receptor selectivity, and we need our version written into the preclinical plan. For Eddie Cherok: the reverse-merger window is not restricted to Phase-2-ready assets, and both of these were in-licensed from abroad in the same week two Republican members of Congress asked FDA to discount Chinese clinical data absent recent site audits. THIRD: the HHS OIG request for information on safe harbors for trial participant remuneration closes today at 5:00 p.m. Eastern; docket re-pulled this morning still shows 8 comments, most recent 2026-08-14, Massive Bio on it and 1104health not. NULLS: nothing from Clasp, Adventris, aSKY, Skylark or Sensorion; zero GPR52 news and still no named buyer for Nxera's NXE'149; nothing new from CAMP4; zero Blackbird mentions. Forward calendar: Ultragenyx UX111 PDUFA September 19, CAMP4 analyst day September 28, the Nxera out-license guided to close this half, and the Clasp GUARDIAN-101 monotherapy readout still guided to this year.

Sourcing Radar — the brake on cochlear hair-cell regeneration finally has a name, and it is still being applied: Angelika Doetzlhofer's lab at Johns Hopkins shows ZBTB20 enforces the mature, non-regenerating state of cochlear supporting cells, that deleting it ACUTELY in already-formed postnatal tissue still reopens the regenerative window, and that the tractable node underneath is a secreted ligand and its receptor — midkine and PTPRZ1 — where adding midkine alone reproduces the phenotype. No competing interests, NIDCD and Rubenstein funding, no company on it: a clean Johns Hopkins license path. And the reason it is a BLACKBIRD lead rather than somebody else's paper is the liability it creates — you cannot broadcast a proliferative axis with oncology history into the ear, you have to restrict it to supporting cells at a controlled level in a semi-enclosed compartment, which is Aletira's entire premise. Frequency Therapeutics ran this thesis on generic stemness pathways and lost the company; this one names the enforcer. Plus a Kennedy Krieger bile-acid paper closed as a WATCH not a build (the rescuing compound is a generic), and a UMB single-molecule protein sequencing platform deferred because bioRxiv has not rendered the body

Sun, 23 Aug 2026 12:00:00 +0000

The Blackbird Brief Sourcing Radar for Sunday, August 23, 2026. LEAD: Morgan CT, Rehman ZU, Doetzlhofer A, "Loss of Zbtb20 disrupts cochlear supporting cell differentiation and maturation and extends the postnatal hair cell regenerative window in mice," bioRxiv, posted 2026-08-20 (Solomon H. Snyder Department of Neuroscience + Otolaryngology-Head & Neck Surgery / Center for Hearing and Balance, Johns Hopkins University School of Medicine; full text read via PDF). Emx2-Cre Zbtb20 knockout delays cell-cycle exit, differentiation and maturation of cochlear supporting cells and leaves the progenitor program (Hmga2, Sox11) switched on at P5 (676 DEGs, 489 up / 187 down, q<0.05). Purified Lfng-GFP+ knockout supporting cells form organoids at roughly 3x the control rate, form larger organoids, proliferate more, and produce significantly more MYO7A+/SOX2+ hair cells per organoid; in Pou4f3-DTR explants they produce more hair cells in both damaged and undamaged tissue. The commercially load-bearing result is the ACUTE deletion: a doxycycline-inducible knockout in already-formed postnatal tissue still increases organoid size and formation rate, so the repression is maintained rather than developmental-only. RNA-seq of the acutely deleted cells (524 DEGs) identifies upregulation of the midkine receptor Ptprz1, and exogenous midkine at 10, 50 and 100 ng/mL reproduces the phenotype (formation-rate benefit at 10 and 50 but not 100 — a real ceiling). IP posture: authors declare no competing interests; funding is NIDCD R01DC019359 and F31DC020882 plus the David M. Rubenstein Fund for Hearing Research; GEO accession GSE342694. Candid limits: organoid and explant only, no in vivo regeneration in mature animals, no ABR/functional hearing recovery, all P5 mouse tissue, and acute deletion drove hair-cell production only in damaged explants. Competitive precedent: Frequency Therapeutics' FX-322 Phase 2b (n=142) missed speech perception and every secondary in 2023; the company discontinued FX-322 and FX-345, cut ~55% of staff, and redirected to remyelination in MS. SECOND ITEM, closed as a WATCH not a build: Opendak M (senior, Kennedy Krieger) and Haughey N (Johns Hopkins, metabolomics), "Bile acid signaling as a therapeutically tractable pathway linking early caregiving adversity to social behavior," bioRxiv, posted 2026-08-19 — social (not non-social) adversity reorganises peripheral bile acid and tryptophan metabolism and basolateral amygdala co-expression networks, and oral chenodeoxycholic acid but not cholic acid rescues the infant affiliative deficit, attributed to CDCA being the most potent endogenous FXR agonist. Clean interests and NIH/Kennedy Krieger funding, but the rescuing compound is a decades-old generic, FXR is crowded and bruised, the endpoint is infant rat social approach, and the responding sex flipped between models. DEFERRED for lack of rendered full text: Taylor JE, Sharma P, Krantz B, "Single-Molecule Proteomics via a Dynamic Translocase and Physics-Informed Machine Learning," University of Maryland Baltimore — anthrax toxin protective antigen as a dynamic nanopore, 98.02% classification accuracy on native clinical peptides including KRAS G12D. Swept and passed: the Johns Hopkins leptomeningeal blood-CSF barrier atlas, the UMB astrocytic Stat3 prion knockout, and a University of Maryland stem-cell secretome proteomics methods paper. No Lieber Institute corresponding-author preprints in the window.

Portfolio Watch — Novartis's venture arm just funded the one company whose Phase 2 tests the exact claim the Lieber GPR52 program is built to make: Kynexis takes a 40 million euro extension (Series A now 97 million euros / ~$110M) led by the Novartis Venture Fund, enrollment COMPLETE on its KAT-II inhibitor KYN-5356 in cognitive impairment in schizophrenia, topline by YEAR END with registrational prep underway — a dedicated, purpose-built, non-dopaminergic cognition readout entering the record before our development candidate is declared, and a positive result and a negative result are each expensive for us in different ways. The franchise is Maryland biology sitting in Naarden because the molecule came from Mitsubishi Tanabe's library and not from Baltimore: the second time in a month, and the transferable warning for the Luetkens cathepsin-B lead. Plus a CORRECTION on Capricor that changes the lesson — FDA's own briefing document says HOPE-3 did NOT meet its pre-specified primary and secondary endpoints and treats the post-study SAP analyses as post hoc, with the final SAP dated one day before unblinding and never agreed with the agency. So this is a case about ANALYSIS-PLAN PROVENANCE, not about whether FDA rewards narrowing, and Genglycos four days early on a pre-agreed surrogate is the contrast case. And a blinded Nature Biotechnology benchmark lands the same day BMS buys AI antibody design: clone ranking was worse than random for every model but one

Sun, 23 Aug 2026 17:00:00 +0000

The Blackbird Brief Portfolio Watch for Sunday, August 23, 2026. LEAD (2026-08-19; lands on the Lieber Institute GPR52 NewCo and on the UM Ventures co-investment surface): KYNEXIS (Naarden, Netherlands) closed a EUR 40M extension to its Series A, taking the round to EUR 97M (~$110M), led by new investor the NOVARTIS VENTURE FUND with Forbion, Ysios Capital and Sunstone Life Science Ventures following on; Novartis's Marianne Uteng joins the board and Mathias Frederiksen as observer. Proceeds fund close-out of the Phase 2 proof-of-concept of KYN-5356 — a purpose-built KAT-II inhibitor licensed from Mitsubishi Tanabe — in cognitive impairment associated with schizophrenia (enrollment COMPLETE, topline guided to end of 2026), preparation for registrational development, and expansion into other cognitive disorders including Alzheimer's. THREE READS. (1) COMPETITIVE CLOCK: the GPR52 value case rests on reaching all three symptom domains, and cognition carries the value. MapLight's cognition signal (effect size ~0.5, covered 08-10) was a subgroup inside a psychosis-powered trial; Kynexis is a trial designed around cognition, in a company built on that claim, with large-pharma venture money and registrational prep. A dedicated non-dopaminergic cognition readout enters the public record before our DC declaration — a positive removes first-in-class, a negative raises the bar on anyone claiming cognitive benefit from any mechanism. (2) UM VENTURES: the kynurenine-pathway biology is Robert Schwarcz's at the University of Maryland School of Medicine (he is a Kynexis scientific co-founder), and the human proof of mechanism came from the Buchanan/Schwarcz tryptophan-challenge study at MPRC — yet the franchise sits in the Netherlands because the composition of matter came from a Japanese pharma library. Directly transferable to the Luetkens cathepsin-B lead on the same UM Ventures x Blackbird surface. (3) BIDDER SET: Novartis was not on our NXE'149 buyer list (Otsuka, BMS/Karuna, Neurocrine, AbbVie/Cerevel) — add them; Nxera still has no named buyer and there was no GPR52 news of any kind in the window. Adjacent: Molecular Psychiatry published a COMT-inhibition-for-cognition review 2026-08-19 co-authored by Lieber's Daniel Weinberger with a first author now at Boehringer Ingelheim, the company that walked away from the GPR52 option in December. SECOND — CORRECTION ON CAPRICOR: the 2026-08-22 PDUFA passed with no announcement from company or agency. This brief has repeatedly said HOPE-3 hit its primary on upper-limb function with the adcomm voting against on a cardiac secondary. FDA's July 27 briefing document says the study "did not meet its pre-specified primary and secondary efficacy endpoints" and that FDA considers analyses based on the post-study SAP versions post hoc and exploratory; final SAP v3.0 is dated 2025-11-24, one day before unblinding, and per the securities complaint (S.D. Cal.; class period 2025-12-17 to 2026-07-26; lead plaintiff deadline 2026-09-28) was never submitted to or agreed with FDA. STAT reported 2026-08-20 that the therapy is heading for a rejection. So this is NOT the test of whether FDA rewards narrowing — it is a case about ANALYSIS-PLAN PROVENANCE, and the operative rule for the SELEXON pre-IND is to agree endpoint and analysis with the division early and in writing. Contrast case, covered in full on 08-20 and used here only as the counterpoint: Ultragenyx's GENGLYCOS accelerated approval four days ahead of its action date on a pre-agreed cornstarch-reduction surrogate with two years of confirmatory data accepted. THIRD — diligence pair for Avi: Bristol Myers Squibb announced an AI antibody discovery collaboration with Chai Discovery on 2026-08-19 (Chai already has Lilly, Pfizer and Novartis), the same day Nature Biotechnology published AIntibody, a blinded prospective benchmark of 511 AI-designed or predicted antibodies from 29 organisations across three tasks — affinity maturation modelled well, clone ranking was worse than random picking for every model but one, out-of-library design was highly variable with many submissions failing to beat standard selections, and success did not transfer across tasks. The rule: ask which of the three jobs a model is doing. FOURTH: the HHS OIG RFI on safe harbors for trial PARTICIPANT remuneration closes Monday 2026-08-24 at 5:00 p.m. ET; docket re-pulled 2026-08-23 still shows 8 comments, most recent 2026-08-14, Massive Bio on it and 1104health not. NULLS: nothing from Clasp, Adventris, aSKY, CAMP4, Skylark or Sensorion; zero Blackbird mentions. Market note for Eddie: Scribe Therapeutics' $150M IPO at $21.50 with a Phase 1 lead asset is up 43% — conservative pricing plus a dated near-term catalyst is what buys an early-stage company a listing right now.

Portfolio Watch — how a pain drug that FAILED a U.S. Phase 3 for futility became worth $500M in eight months: Ambros takes Werewolf's Nasdaq shell with a $150M oversubscribed PIPE (RA Capital + Janus Henderson) for neridronate in CRPS-1 — and the protocols show the molecule, the 400 mg cumulative IV dose over 10 days, and the Week-12 pain-intensity primary are all UNCHANGED from the Grünenthal trials stopped for futility in 2019. Everything that changed is the entry door: warm subtype only, ≤6 months from onset, positive centrally-read triple-phase bone scan, CRPS-2 and CRPS-NOS excluded, Pain Catastrophizing Scale ≥40 excluded, eDiary run-in. That is what staking Crohn's PAIN axis actually costs for the Slusher gut-restricted GCPII program — a stratification package, not a novel mechanism. Plus the second reverse merger in eight days prices the shell template for Eddie (Werewolf $47.5M, legacy holders 6.8%), and a conditionally-activated-cytokine PLATFORM company becoming a listing vehicle is the clarifying read for Clasp. Today is Capricor's PDUFA — the live test of whether an agency that rewards NARROWING really does — with an activist demanding board seats the day before. And the HuidaGene death mechanism is now stated in public as complement and cytokine activation after HIGH-DOSE SYSTEMIC AAV, which is our own durable line in someone else's words

Sat, 22 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Saturday, August 22, 2026. LEAD (2026-08-21; lands on the JHU gut-restricted GCPII program, on Eddie's exit-template file, and on Clasp): WEREWOLF THERAPEUTICS (Nasdaq: HOWL) and AMBROS THERAPEUTICS announced an all-stock merger plus a concurrent OVERSUBSCRIBED $150M private placement co-led by RA CAPITAL MANAGEMENT and JANUS HENDERSON. Combined company renames to Ambros, San Diego, ticker AMBX; Ambros valued at $500M, Werewolf at $47.5M; ownership ~71.7% Ambros / ~21.5% new investors / ~6.8% legacy Werewolf; funded through 2028 topline and into 1H 2029; CEO Jay Hagan. On top of a $125M Series A closed 2025-12-16 (RA Capital + Enavate Sciences co-led, with Abiogen Pharma, Janus Henderson, ARKIN BIO — note: also Skylark Bio's investor — Balyasny, Transhuman, Adage) = ~$275M in eight months. THE ASSET is NERIDRONATE, an amino-bisphosphonate licensed from Abiogen Pharma, approved in Italy for CRPS-1 since 2014 and given to ~600,000 patients across its indications there; CRPS-1 is an orphan pain disorder, ~65,000 new U.S. cases/yr, NO FDA-approved treatment; Breakthrough Therapy + Fast Track + Orphan Drug in hand. THE ANALYTICAL CORE, from putting both registry protocols side by side: GRÜNENTHAL ALREADY RAN THIS MOLECULE INTO A U.S. REGISTRATIONAL PROGRAM AND FAILED. Its two Phase 3s (KF7013-02 / NCT03530345, n=182; KF7013-04 / NCT03560986, n=267) were TERMINATED in 2019 after a pre-planned POOLED INTERIM FUTILITY analysis — the futility criterion was an observed neridronic acid vs placebo mean difference ≥ −0.3 points on the 11-point NRS, and an independent statistician recommended stopping. Ambros's CRPS-RISE (NCT07210515, n=270, 24 U.S. sites, recruiting since 2026-04-14, topline 2028) keeps EVERYTHING that failed: SAME molecule, SAME 400 mg cumulative IV dose across Days 1/4/7/10, SAME primary endpoint (change from baseline in average pain intensity, 11-point NRS, Week 12). WHAT CHANGED IS ENTIRELY THE ENTRY DOOR. Grünenthal: CRPS broadly per Budapest criteria, duration up to TWO YEARS since onset, no subtype restriction, no imaging requirement, and prior failure of ≥2 treatments REQUIRED (i.e. chronic, refractory, heterogeneous), 71 sites across 8 countries. Ambros: CRPS-1 only, single limb, no known peripheral nerve injury, ≤6 MONTHS since onset, WARM SUBTYPE ONLY (edema plus ≥2 of obvious redness, ≥1°C temperature rise, moderate-to-severe edema), POSITIVE TRIPLE-PHASE BONE SCAN with a CENTRAL READ, CRPS-2 and CRPS-NOS excluded, cases with no identified inciting event excluded, PAIN CATASTROPHIZING SCALE ≥40 EXCLUDED, and an eDiary-compliance gate during screening. Five enrichment gates on an unchanged drug at an unchanged dose against an unchanged endpoint — and a syndicate that knew about the first failure paid $500M for the second attempt. READ (1) — FOR THE GUT PROGRAM (Hemaka + Barbara Slusher's JHU Drug Discovery team). Since the 2026-08-14 Fibrx transaction the durable line here has been that inflammation and fibrosis in Crohn's are spoken for and PAIN IS THE LAST UNSTAKED AXIS. This deal prices what staking it costs, and the currency is NOT mechanism — nobody paid for novelty; it is a decades-old bisphosphonate. They paid for a STRATIFICATION PACKAGE. So the program question is not whether gut-restricted GCPII inhibition relieves visceral pain in a mouse; it is WHICH IBD PATIENTS, IDENTIFIED BY WHAT OBJECTIVE MEASURE, BEFORE DOSING. If there is an enrichable subpopulation (a visceral-hypersensitivity phenotype, an imaging or biomarker readout, a disease-duration window), that is the registration path and it belongs in the program plan NOW, not after a Phase 2 misses on all-comers. CANDID LIMITS: neridronate acts on bone resorption and has nothing to do with the enteric nervous system, and an orphan indication with no approved therapy is a far cleaner regulatory setup than pain in Crohn's, where every patient is already on an approved anti-inflammatory. What transfers is trial architecture and financing, not biology. READ (2) — FOR EDDIE. Second reverse merger into a Nasdaq shell in EIGHT DAYS (after Skye + Redx → Fibrx on 08-14), in a year with roughly 18 of them, the most since 2018 — and this one PRICES the template: the shell went for $47.5M and legacy holders kept 6.8%. For a Phase-2-ready Baltimore asset that does not want a conventional offering, that arithmetic is now visible on two deals in one week. READ (3) — THE UNCOMFORTABLE HALF, FOR CLASP. Werewolf was a PLATFORM company: conditionally activated IL-2 (WTX-124) and IL-12 (WTX-330) INDUKINE molecules — precision delivery in oncology, a first cousin to how several portfolio assets describe themselves. It opened a strategic review 2026-02-24, could not get a clinical asset over the line, and ended as a listing vehicle worth 6.8% of the company that took its ticker. Not alarming, clarifying: the platform story is not what is funded right now; the GUARDIAN-101 monotherapy readout is. SECOND ITEM — CALENDAR, ALETIRA-RELEVANT: TODAY (2026-08-22) IS CAPRICOR'S PDUFA TARGET for deramiocel in Duchenne. The 07-29 adcomm voted 9–3 against, but on CARDIAC FUNCTION (a secondary), not on upper-limb function (the primary the trial hit); the company is amending its BLA with 24-month HOPE-3 extension data and narrowing to the endpoint it hit, which may extend the date. This is the live test of the 08-18 framing (an agency RIGOROUS ON EVIDENCE, FLEXIBLE ON PATH, that rewards NARROWING) — the frame sitting underneath the SELEXON expression-control pitch. Ultragenyx printed four days early this week, so a weekend action date does not mean nothing happens; watch today and Monday. BURN NOTE (Matt + Emily): on 08-21 KAOS CAPITAL went public demanding an immediate meeting, two independent director nominations, a board-led M&A / Strategic Alternatives Committee, and a capital-preservation plan — the day before the action date. Capricor cash $318.1M (YE2025) → $237.9M (2026-06-30): $80.2M in six months while awaiting a decision on a filed BLA. THIRD — GENE-THERAPY CLIMATE: Reps. JOHN MOOLENAAR (R-MI) and BEN CLINE (R-VA) formally asked FDA to reject Chinese clinical data unless the trial sites have had a recent FDA audit, and to review products already approved on it. Trigger is three deaths: HuidaGene HG302 (DMD) and the Shanghai Jiao Tong six-year-old — both aired here 08-07 — plus one NEW to this brief, a systemic sclerosis patient who died after RiboX's in vivo CAR-T, disclosed 2026-08-09 after FDA had cleared a U.S. IND in that program. THE USABLE LINE: the HuidaGene death is now described in public as ACUTE RESPIRATORY DISTRESS SYNDROME IN THE SETTING OF SEVERE COMPLEMENT AND CYTOKINE ACTIVATION FOLLOWING HIGH-DOSE SYSTEMIC ADMINISTRATION OF AN AAV VECTOR — the cleanest external statement yet of the sentence Hugh Wells and Avi were asked on 08-20 to tighten: THE IMMUNE PROBLEM IN SYSTEMIC AAV IS THE CAPSID AND THE DOSE, NOT THE CASSETTE. Do not let a pitch imply SELEXON would have prevented it. It DOES support the beachhead: a local micro-dose into a small semi-enclosed compartment does not pay that tax. Secondary read for Eddie: policy is moving toward DATA PROVENANCE, so an in-licensing package leaning on Chinese investigator-initiated data is now a diligence line item. FOURTH — 1104HEALTH, WITH A CORRECTION TO YESTERDAY: the 08-21 episode said Friday was the last business day to file on HHS OIG docket HHSIG-2026-0034. WRONG — comments close MONDAY 2026-08-24 AT 5:00 P.M. ET, so there is ONE WORKING MORNING LEFT, not zero. Docket re-pulled 2026-08-22 via the regulations.gov v4 API: STILL EIGHT comments, most recent still 2026-08-14; Massive Bio (a listed 1104health competitor) still on it, 1104health still not. NULLS / SHORT LINES: no GPR52 news of any kind and Nxera still has no named buyer for NXE'149; Otsuka scrapped an early psychedelic trial from its Mindset acquisition — flagged only because Otsuka is on the specialty-neuro bidder short list for NXE'149, and one discontinued early trial is thin evidence, so hold it loosely; nothing from CAMP4 since the 08-13 Q2 report (Sept 28 analyst day + Q4 first dosing stand); no fresh mutant-KRAS catalyst for aSKY; no Clasp or Adventris update; nothing from Skylark or Sensorion; Ultragenyx's UX111 Sanfilippo decision still 2026-09-19; ZERO in-window mentions of Blackbird Laboratories, BioVentures or the BioHub.

Portfolio Watch — Wednesday's Skylark device question, answered from the only approved label in our own lead indication: Regeneron's OTARMENI goes in through a CATALOG CATHETER (a Vygon Premicath neonatal PICC, 1 Fr / 28G, in a required kit of BD needles and syringes) placed through the round window with electrode insertion forceps after a standard mastoidectomy — and CHORD registered DB-OTO as a plain genetic intervention with ZERO device endpoints across 29 secondaries, while Skylark (SKY-CAT), Akouos/Lilly (two device configurations) and Sensorion (three device secondaries) all registered COMBINATION PRODUCTS. Three of four took device risk; the one that got approved declined it, so SKY-CAT is a bet Skylark made, not a moat we must match — but two funded groups engineered a catheter anyway, which sharpens the real question: elective for an otoferlin-shaped target, necessary for a connexin-shaped one? The same label also puts EXPRESSION RESTRICTION IN THE COCHLEA on an approved product four months before Ultragenyx — a dual AAV1 with "an engineered hair cell-specific promoter derived from regulatory elements of myosin 15" — a better citation for Geoff than yesterday's liver drug, and an uncomfortable one: the principle is settled, and it was settled by a PROMOTER, so SELEXON's claim has to move from ADDRESS to LEVEL. Plus the dual vector is the cassette-space tell. Second: Merck buys community-oncology enrollment from Sarah Cannon's Accelero — 19% enrollment vs a 7% national average, a published bar 1104health must approach in a network it does not own — and a CORRECTION on the OIG docket (it is about paying PATIENTS, not physicians) plus the finding that matters more: the eight commenters include MASSIVE BIO, our own listed competitor, and 1104health is not among them with one business day left

Fri, 21 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Friday, August 21, 2026. A quiet tape, so the day is ORIGINAL WORK that closes the action item opened Wednesday: what is Aletira's delivery answer in the ear, and is it someone else's device whose label someone else will own. LEAD (lands on flagship ALETIRA, three ways, all from primary documents): read OTARMENI (lunsotogene parvec-cwha, Regeneron) — approved 2026-04-23, the only approved gene therapy for genetic hearing loss in the world. (1) THE DELIVERY ANSWER, AND IT IS BETTER THAN WE FEARED. The label requires an Administration Kit, and the kit contains a BD PrecisionGlide 21G needle, a BD 1 mL syringe, a BD PlastiPak 3 mL syringe, a Luer-Lok tip cap, and ONE VYGON PREMICATH CATHETER — 1 French, 28 gauge, 0.33 mm: a commercially available neonatal PICC line designed for peripheral venous access in premature infants under one kilogram, orderable from a medical supply distributor. The procedure in the label is a standard mastoidectomy, an opening of the facial recess, and placing the catheter tip through the round window membrane USING ELECTRODE INSERTION FORCEPS — i.e. cochlear implant surgery with cochlear implant instruments. Be precise: the kit is required and is a regulated co-packaged component, but every part in it is a catalog part. REGENERON DID NOT BUILD A DEVICE. Now the registrations, all four pulled off ClinicalTrials.gov this morning: Skylark's SONIX registers SKY-GJB2 as a COMBINATION_PRODUCT with the SKY-CAT named inside the PRIMARY safety endpoint and again as its own secondary; Akouos (now Lilly) registers AK-OTOF-101 as a COMBINATION_PRODUCT with TWO device configurations in the same trial ("Akouos Delivery Device" and "Akouos Delivery Device and Precision Delivery Mechanism") plus a device-performance secondary; Sensorion's AUDIOGENE registers SENS-501 as a COMBINATION_PRODUCT with THREE device secondaries (clinical performance, safety, usability of the administration system); and Regeneron registered DB-OTO in CHORD as a plain GENETIC intervention with 29 secondary outcomes and NOT ONE device endpoint. THREE OF FOUR PROGRAMS TOOK DEVICE DEVELOPMENT RISK; THE ONE THAT REACHED APPROVAL DECLINED IT. REFRAME of Wednesday's read: SKY-CAT is not a moat Aletira must match, it is a REGULATORY SURFACE Skylark chose to add — in a combination product a device deficiency can hold up your drug, and Skylark wrote that exposure into its own primary endpoint — while our delivery answer already exists on an approved label and can be ordered from a catalog. TWO LIMITS, and the second is the useful one: (a) a catalog catheter is available to everybody, so this NEUTRALIZES their advantage rather than creating ours; (b) two independent well-funded groups engineered a delivery device anyway, and that is evidence, not vanity — the likeliest reason is the TARGET. Regeneron needed inner hair cells; Skylark's target and Aletira's harder targets need broad, even coverage of SUPPORTING cells along the length of the cochlear duct. SHARPER QUESTION FOR AVI + HUGH WELLS: is an engineered device ELECTIVE for an otoferlin-shaped target and NECESSARY for a connexin-shaped one? If the latter, we need a delivery PARTNER rather than a delivery program — and there are now at least three device designs in the clinic to partner with instead of build. (2) THE SAME LABEL PUTS EXPRESSION RESTRICTION IN THE COCHLEA ON AN APPROVED PRODUCT, FOUR MONTHS BEFORE ULTRAGENYX. The construct is a DUAL AAV serotype 1 carrying, in the label's own words, "an engineered hair cell-specific promoter derived from regulatory elements of myosin 15 (Myo15)" driving OTOF cDNA. So yesterday's ladder now has a rung IN OUR OWN INDICATION — same tissue, same modality, same agency — a better slide for Geoff Lynn than the liver drug, and he should have both. Be exact about the rung: ENGINEERED, but DERIVED FROM Myo15 regulatory elements, i.e. between a synthetic tissue-specific promoter and a fully native one. THE UNCOMFORTABLE HALF: restricting expression in the cochlea is now claimed at the APPROVAL level and it was claimed BY A PROMOTER, so our argument cannot be about the principle — the principle is settled and somebody else settled it. It has to be about what a promoter cannot do. A PROMOTER IS AN ADDRESS; it tells the cassette which door to go in. Connexin 26 must land in supporting cells at roughly the right LEVEL, because gap junctions are dose-sensitive and overexpressing a connexin in the CORRECT cell type is its own toxicity. AN ADDRESS DOES NOT GIVE YOU A LEVEL. That is the claim SELEXON should be making — narrower, more technical, far more defensible than "we control expression." (3) THE CASSETTE TELL, on Wednesday's open diligence item: OTARMENI is a DUAL VECTOR because the otoferlin coding sequence does not fit in one AAV, so Regeneron split it and relies on the halves reassembling in the cell. That is the approved workaround for running out of cassette space and it costs two vectors' worth of capsid load — in a year when yesterday's Ultragenyx label is a fresh reminder of what capsid load buys you. When the Series A question arrives about what a selective exon costs in kilobases, the answer we do not want to give is "we would go dual." SECOND ITEM, on 1104HEALTH, in two parts. (a) MERCK SIGNED SARAH CANNON RESEARCH INSTITUTE (press release 2026-08-13, trade coverage 08-19) to run oncology trials through SCRI's Accelero delivery model at community sites: ~1,500 oncology physicians, 200+ locations, 20+ states, 900+ first-in-human trials to date. Published metrics: site activation 50% FASTER, 95% FEWER data changes (EHR-to-EDC transfer instead of coordinator re-entry), and ENROLLMENT AT 19% AGAINST A 7% NATIONAL AVERAGE. THE VALIDATION IS REAL and comes from the largest oncology sponsor there is — community enrollment is the bottleneck and it is worth paying to fix, which is Rose Wang's market thesis with Merck's signature on it. THE CANDID HALF: look at HOW Merck bought it — not a marketplace but a single VERTICALLY INTEGRATED network that owns its sites and sells one operating model with the data plumbing built in; and look at the mechanism they credit — SCRI REMOVED the coordinator's work, while 1104health PAYS FOR the coordinator's work. Same bottleneck, opposite solution, and Merck bought the other one. 19-against-7 is now a PUBLISHED BAR that 1104health must approach in a network it does not own. Question for Eddie + Rose: is the buyer the sponsor who CANNOT build a captive network (everyone smaller than Merck), and does the pitch lead with the economics of not having to own the sites? (b) THE OIG DOCKET, WITH A CORRECTION WE OWE THE RECORD: this brief has been describing HHSIG-2026-0034 / FR Doc 2026-12676 as a rulemaking about paying physicians for trial-coordination work. Reading the notice itself, IT IS NOT. All FOURTEEN enumerated questions concern remuneration TO CLINICAL TRIAL PARTICIPANTS — patients: cost-sharing subsidies, transportation, lodging, parking, childcare, meals, stipends, compensation for time, completion incentives. There is nothing in it about paying a practice for coordination burden, so the direct line to 1104health's mechanism is WEAKER than we have been implying. WHAT IS ACTUALLY WORTH SOMETHING (pulled the commenter identities this morning, not just the count): eight comments, and while it remains true that none is from a trade association or a major sponsor, the list is precisely 1104health's technology category — MASSIVE BIO, INC. (filed 07-30; a company already on our own competitive watch list for 1104health), MURAL HEALTH TECHNOLOGIES (08-10; a participant-payment platform), FLORENCE HEALTHCARE (07-30; site software, comments "limited to the operational questions"), the ONCOLOGY NURSING SOCIETY (07-30), THE STARR COALITION (07-01), TechInHSR LLC (07-01), and an individual-capacity submission (08-14) addressing Questions 2, 4, 6, 10, 11, 12 and 14. 1104health is not on the list. Two questions do bear on a platform: Q5 asks WHO should be permitted to provide participant remuneration and enumerates "the clinical trial sponsor, investigators, other providers or suppliers"; Q11/Q12 ask what ARRANGEMENTS NECESSARY TO EFFECTUATE that remuneration should be protected. Those are where an intermediary asks not to be written out of a safe harbor before it is drafted. COMMENTS CLOSE MONDAY 2026-08-24 AT 5 P.M. ET — TODAY IS THE LAST BUSINESS DAY, and a direct competitor filed three weeks ago. SHORT LINES: Merck/Moderna intismeran follow-on — ~$45B added to Moderna's market cap and analysts asking whether the vaccine can be used as broadly as KEYTRUDA; no new data, and it still argues the off-the-shelf version is worth MORE for Clasp and Adventris. Boehringer Ingelheim placed a Phase 1 of an UNDISCLOSED small molecule on hold (08-20) over an UNDISCLOSED safety event; Boehringer is the party that walked away from its option on Nxera's GPR52 agonist in December 2025, so it is flagged — but the program is not named and Fierce article bodies were unreachable from this host today, so NO inference is drawn. Amgen walked from its $500M+ Crohn's collaboration with TScan and TScan says its pipeline strategy is unchanged (inflammation-axis datapoint, no move on the Slusher gut-restricted GCPII program). Ultragenyx's second AAV decision, UX111 for Sanfilippo type A, is confirmed for 2026-09-19. NULLS: no GPR52 news of any kind and Nxera still has no named buyer for NXE'149; nothing from CAMP4 (Sept 28 analyst day remains the event); no fresh KRAS catalyst for aSKY; no Clasp or Adventris update; nothing from Skylark or Sensorion since our own Wednesday work; ZERO in-window mentions of Blackbird Laboratories, BioVentures or the BioHub. COVERAGE-GAP ITEM, explicitly NOT news and nine months old: a Broad Institute preprint (Fei Chen lab, Todd Golub on the author list) describes SPICE, a generative model that DESIGNS cell-type-specific RNA splicing elements de novo — an MPRA across 46,372 human exons, a transformer-based generator, designed elements validated in dual-colour frameshift reporters across 43 cell lines; no AAV, no in vivo, and A PATENT APPLICATION HAS BEEN FILED. On Wednesday we said the splicing layer was untaken; clinically that holds, on the IP it does not. This belongs in the same envelope as the freedom-to-operate work on Skylark's 3'UTR claims, and since Fei Chen has founded two companies and sits on the scientific advisory board of an RNA-editing company, assume it becomes one.

Portfolio Watch — the FDA approves a gene therapy that regulates its own transgene: Ultragenyx's GENGLYCOS (AAV8, G6Pase under the gene's OWN NATIVE PROMOTER, so the hepatocyte still answers to insulin and cortisol) is the best external endorsement Aletira's expression-control thesis has ever had — and its label is also the sharpest correction to it (71% transaminase elevations, corticosteroids for ALL patients, 24% adrenal insufficiency, 10% anaphylaxis: expression control is not an immunology answer). Plus the post-marketing CONTROL ARM is patients excluded by anti-AAV8 antibodies (a systemic-delivery tax the ear largely avoids), and the approval carries a rare pediatric PRIORITY REVIEW VOUCHER on a program that sunsets 2029-09-30 while vouchers trade at $180-200M — arithmetic that argues for licensing SELEXON into someone else's IND. Second: Merck/Moderna's INTerpath-001 is the first Phase 3 win for an individualized neoantigen therapy, validating the premise under Adventris and Clasp while pricing the bespoke route and making SHARED-neoantigen, off-the-shelf approaches worth more

Thu, 20 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Thursday, August 20, 2026. LEAD (2026-08-19; lands on flagship ALETIRA, four ways): FDA granted ACCELERATED APPROVAL to Ultragenyx's GENGLYCOS (pariglasgene brecaparvovec-opnr, formerly DTX401) for glycogen storage disease type Ia in patients 8+. Single IV infusion, AAV8 capsid, G6Pase carried under the gene's OWN NATIVE PROMOTER — chosen deliberately so treated hepatocytes still respond to insulin and cortisol. (1) TAILWIND / CITATION: approved AAV products already use restricted promoters (Hemgenix, Roctavian in liver; Elevidys in muscle), but this is native-promoter PHYSIOLOGICAL regulation, not just cell-type restriction. Stack it against yesterday's finding that Skylark restricts at the 3'UTR and Aletira restricts at the splice: three companies, three elements on the same molecule, one shared premise — endogenous regulatory grammar should govern the transgene — and that premise now has an FDA approval behind it. Geoff Lynn should be quoting it within the week. (2) THE CORRECTION, on the record before an investor says it to us: native-promoter control did NOT make the product safe. Label shows 71% ALT/AST elevations, corticosteroids administered to ALL patients as standard mitigation, 24% adrenal insufficiency (including serious events), 14% Cushingoid features, 10% anaphylaxis, contraindication in severe hepatic fibrosis/cirrhosis, plus vector-integration/tumorigenicity language. The immune problem in systemic AAV is the CAPSID AND THE DOSE, not the cassette. Pitch SELEXON as a general safety platform and this label loses us the room; the defensible claim is narrower and stronger — SELEXON solves OFF-TARGET EXPRESSION, a different problem from capsid immunogenicity. (3) BEACHHEAD ARGUMENT: the agreed post-marketing control group is patients who sought commercial treatment but CANNOT BE DOSED because they carry anti-AAV8 antibodies — seroprevalence exclusion is now reliable enough to serve as a comparator arm. That is a systemic-delivery tax; a local micro-dose into a small semi-enclosed compartment largely avoids it. (4) MONEY AND CLOCK: the approval carried a rare pediatric disease PRIORITY REVIEW VOUCHER. Congress revived the lapsed program in the February 2026 appropriations act; FDA cannot award vouchers under it after 2029-09-30, with no separate designation deadline, so the binding constraint is the APPROVAL date. 2026 voucher sales: Rocket $180M, Denali $195M, Immedica $200M, Jazz $200M. Aletira is preclinical — a wholly owned program cannot reach approval by September 2029; Skylark (dosing since May) and Sensorion (CTAs filed in Canada and France) plausibly can. Read: the voucher is not ours on a go-it-alone path, but it makes a bolt-on expression layer worth more TODAY to a partner already in the clinic — an explicit argument for licensing SELEXON into someone else's IND in parallel with our own program. Ultragenyx has a second AAV decision (Sanfilippo type A) dated September 19. SECOND ITEM (2026-08-19; lands on the KRAS/immuno-oncology board — Adventris in the BioHub, Clasp where BioVentures is in the Series A): Merck and Moderna's Phase 3 INTerpath-001 met its primary endpoint of recurrence-free survival and key secondary of distant metastasis-free survival — 1,137 patients, completely resected stage IIB-IV melanoma, individualized neoantigen therapy (intismeran autogene, up to 34 neoantigens, 9 doses) plus KEYTRUDA vs KEYTRUDA alone, no new safety signals; Moderna stock nearly doubled. First positive Phase 3 for an individualized neoantigen therapy and for an mRNA cancer therapy. The validated premise — teach T cells to see tumor-specific mutations and outcomes improve — is the same premise under Adventris's mutant-KRAS peptide vaccine and Clasp's pHLA-directed T-cell engagers. Candid read: the result simultaneously PRICES the bespoke route in public, which makes SHARED-neoantigen, off-the-shelf versions (six KRAS mutations in one vaccine; an engager made once) MORE valuable, not less — the sentence Adventris and Clasp should be using in partnering conversations this quarter. Caveats: no hazard ratios disclosed (deferred to a medical meeting), interim analysis with OS still running, and adjuvant melanoma on pembrolizumab is the friendliest possible setting; the other eight INTerpath trials tell us whether it travels. THIRD: 1104health / HHS OIG docket HHSIG-2026-0034 re-pulled this morning — STILL eight comments, most recent 2026-08-14, still not one from a trade association or major sponsor; comments close Monday 2026-08-24 at 5 p.m. ET, two business days after today. SHORT LINES: Teva's anti-TL1A duvakitug RELIEVE UCCD Phase 2b published 08-18 (topline known since 2024) — one more brick in the inflammation axis, reinforcing that PAIN is the open axis for the Slusher gut-restricted GCPII program; no GPR52 news and Nxera still has no buyer for NXE'149; CAMP4 nothing since the 08-13 quarter (Sept 28 remains the event); no fresh KRAS catalyst for aSKY; no Clasp update; nothing new from Skylark or Sensorion (docket watch on the Skylark assignee stands); zero in-window Blackbird mentions.

Portfolio Watch — yesterday's Skylark action item, closed: the SONIX protocol and Skylark's patent estate answer the question that was going to be asked in Aletira's Series A. Their expression restriction sits at the THREE-PRIME UTR, not at the splice (the promoters in the filing are ordinary ubiquitous ones) — so SELEXON's layer is still ours, but the neighbor is nearer than expected and needs a real freedom-to-operate read; the protocol is also smaller and softer than the press release (n=10, safety-only primary, ABR secondary, opened May 28); and Skylark has a PROPRIETARY DELIVERY DEVICE (SKY-CAT) with its own endpoint, a moat we have no answer to. Plus $125M creates Fibrx around a GI-RESTRICTED oral pan-ROCK inhibitor for fibrostenotic Crohn's — validating gut restriction as an architecture while claiming the fibrosis axis, leaving PAIN as the last open ground for the Slusher GCPII program; and the NEJM Elevidys case where vector arrived at high copy number and protein didn't, which sharpens our expression argument and creates a cassette-length diligence question

Wed, 19 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Wednesday, August 19, 2026. Not a headline day — an ANSWER day. LEAD (original work; closes the action item issued in the 08-18 episode; lands on flagship ALETIRA): pulled the SONIX protocol (ClinicalTrials.gov) and Skylark Bio's patent estate. THREE FINDINGS. (1) THE MECHANISM — Skylark still has NOT disclosed it in words; all company language (including the first-patient-dosed boilerplate) says only "a proprietary capsid engineered for broad transduction" enabling "targeted, micro-dose delivery with cell-specific expression control." But there is exactly ONE published Skylark application worldwide: a recombinant AAV filing, priority 2024-02-20, filed 2025-02-20, published 2025-08-28, assignee Skylark Bio Inc., inventors May Kwang-Mei Wang, William J. Neidermyer Jr., Will J. McLean. It claims an ENGINEERED THREE-PRIME UNTRANSLATED REGION built from "two or more non-identical nucleotide sequences" derived from the target gene's own native 3'UTR, to promote tissue- and cell-type-specific expression and confine AAV-driven expression to cells with endogenous expression of that gene. SO: NOT SPLICING — SELEXON's layer is untaken and Geoff Lynn can say so. BUT closer than expected: the promoters listed in the filing are the ordinary ubiquitous ones (CBA, CMV, CAG, SV40, EF1-alpha), so Skylark is NOT restricting at the transcriptional door at all — it restricts POST-transcriptionally, at the other end of the same molecule, using the native gene's own regulatory grammar. Same instinct as SELEXON, one element over. ACTION UPGRADED: this is no longer nice-to-know but a real FREEDOM-TO-OPERATE opinion from counsel BEFORE the raise, specifically on the 3'UTR claims and how broadly they are drafted. WRINKLE that sets the follow-up: the published filing covers the PENDRIN program (SLC26A4 — Pendred syndrome, DFNB4, enlarged vestibular aqueduct; named target cells root cells, spindle cells, outer sulcus cells, spiral prominence epithelium, endolymphatic sac, vestibular cells; capsids AAV1/2/5/7/8/9, AAV-PHP.B, AAV-S preferred), NOT GJB2. We have seen the method, not the lead-program claims — assume a parallel GJB2 filing is pending and set a DOCKET WATCH on the assignee. (2) THE PROTOCOL IS SMALLER AND SOFTER THAN THE PRESS RELEASE — enrollment TEN subjects; study start 2026-05-28 (so "first patient dosed" on 08-11 came ~10 weeks after the trial opened); primary completion Dec 2027, completion Jun 2028; PRIMARY ENDPOINT IS SAFETY ONLY (AEs from the therapy AND from the administration procedure/device); change in ABR threshold is SECONDARY. So the year-end 2026 "preliminary data" is safety plus a handful of hearing thresholds from a ten-child open-label study — a real first-in-human signal, NOT an efficacy verdict. Do not let it be treated as one this fall in either direction, and do not over-brace for it. Eligibility narrows the beachhead further: biallelic pathogenic/likely-pathogenic GJB2 variants, bilateral SNHL at 85+ dB HL at one or more frequencies 500-4000 Hz in the treated ear, ages 9 months to 7 years, measurable ABR response required, autosomal-dominant GJB2 and bilateral cochlear implants excluded — i.e. severe-to-profound RECESSIVE PEDIATRIC disease, while Sensorion claims three populations for SENS-601 including adults with early-onset presbycusis. GJB2 will FRAGMENT by severity and age rather than be won in one shot; real ground remains. Sites: Massachusetts Eye and Ear (Boston — the institution of Skylark advisor Daniel Lee), Hearts for Hearing Oklahoma City (Rene Gifford, PhD), Lehigh Valley Hospital-Cedar Crest (Ravi N. Samy, MD FACS). (3) THERE IS A DEVICE AND WE HAVE NO ANSWER TO IT — the protocol names the SKY-CAT, a proprietary intracochlear delivery catheter, carrying its OWN secondary endpoint (device safety and effectiveness in delivering the therapy, one month post-op). This is a COMBINATION PRODUCT: Skylark will own clinical data on the cassette AND on the surgical access route. In the ear, delivery may prove as defensible as the payload, and it is a moat with nothing to do with expression control. Question for Geoff Lynn and Hugh Wells: what is our delivery answer in the ear, and is it someone else's device whose label someone else will own. COMPANY CALIBRATION (fills a gap in our file, which had investors as undisclosed): the investor is ARKIN (Arkin Bio / Arkin Capital, Israel), position described as "NOTE to Series A" — so the company that just dosed a child is early-stage financed, not a juggernaut. Arkin's own program language: "targeted micro-dose... potentially first- and best-in-class." Note the substitution: their safety pitch leans on giving LESS VIRUS, a partial substitute for restricting expression, and that is the argument we will be positioned against. Inventor Will McLean co-founded Frequency Therapeutics (MIT TR35, 2018); Neidermyer built the AAV platform at Ally Therapeutics (PhD virology, Harvard). Real inner-ear people, modest money, moving fast: we are not outgunned on capital, we are outpaced on the clock. SECOND ITEM (2026-08-14; lands on the JHU/SLUSHER GUT-RESTRICTED GCPII PROGRAM): Skye Bioscience and Redx Pharma agreed to combine into FIBRX THERAPEUTICS (Nasdaq; close expected Q4 2026) with $125M of new money — a ~$68M PIPE plus a $36M Series A led by Abingworth with British Business Bank, NEXTBio Capital, 5AM Ventures and Redmile, plus an equity line up to $22M; Redx management runs it (CEO Lisa Anson, ex-AstraZeneca; CMO Mei-Lun Wang), funded into 2029. Lead asset RXC008: oral, GI-RESTRICTED pan-ROCK inhibitor for FIBROSTENOTIC CROHN'S (the stricturing axis, where the answer today is surgery and nothing approved touches the fibrosis); FDA Fast Track Jan 2026, open IND, Phase 2 starting H2 2026, topline H2 2028. Pipeline also zelasudil (RXC007, selective ROCK2, finishing Phase 2 in IPF) and a preclinical DDR program (IND 2027). CUTS TWO WAYS. VALIDATING: RXC008's Phase 1 showed favorable tolerability with gut TISSUE EXPOSURE and NO clinically relevant systemic breakthrough and no hypotension — precisely the clinical-pharmacology package our gut-restricted GCPII program has to produce, now funded at $125M as a fundable, de-riskable architecture in Crohn's specifically. Cite it in Hemaka's deck. NARROWING (the durable line): Crohn's is being staked axis by axis — inflammation is thoroughly spoken for (TL1A, IL-23, integrin, S1P, JAK), and as of Friday FIBROSIS is spoken for too, with money and Fast Track behind it. NOTHING in the Fibrx package touches PAIN. So the standing instruction — differentiate on the enteric-nervous-system pain axis or the positioning is not defensible — is no longer merely true, it now describes the LAST OPEN AXIS in the indication. Build the positioning on that sentence and stop hedging it. STRUCTURAL FOOTNOTE for Eddie: the route to Nasdaq here is a reverse merger into a shell with contingent value rights paying legacy holders out of the old assets — 17 such deals in 2026, the most since 2018, and a live exit template for a Phase-2-ready Baltimore asset that does not want a conventional offering. THIRD ITEM (NEJM letter, Samelson-Jones et al., Children's Hospital of Philadelphia, NEJM 2026;395(6):615-617, published late July; the MOLECULAR detail was worked through by the neurology trade press on 2026-08-18; ALETIRA thesis + a diligence question): a 17-year-old with DMD treated with Sarepta's Elevidys (delandistrogene moxeparvovec) developed myocarditis with symptomatic atrial fibrillation and mild thrombotic microangiopathy 2 days after infusion, then recurrent troponin I elevations and worsening cardiac function around day 42, treated with IV methylprednisolone. THE MOLECULAR FINDING IS THE POINT: cardiac tissue had HIGHER vector copy number than skeletal muscle — the vector ARRIVED, in quantity — yet immunofluorescence found microdystrophin in only ~5% of cardiac fibers and ~10% of skeletal fibers, and Western blot put microdystrophin BELOW 1% of normal dystrophin in BOTH tissues. Transcripts were unevenly distributed across the cassette, far more from the 5-prime end than the 3-prime end, which the authors read as INCOMPLETE PACKAGING of the transgene cassette into the capsid. Twelve myocarditis cases after this product appear in WHO VigiAccess and FDA FAERS; authors conclude the case "highlights the need for safety monitoring beyond that recommended on the label." SUPPORTING READ (deeper than our current pitch): a field that scores itself on biodistribution just published a case where DELIVERY SUCCEEDED AND EXPRESSION FAILED. We have been arguing that off-target expression causes toxicity; this licenses the more fundamental argument that you do not control expression merely because you controlled delivery. COUNTER-READ (the reason to brief it): the mechanism here is CASSETTE LENGTH and PACKAGING INTEGRITY — a cassette pushed to the AAV packaging limit yields truncated genomes, truncated genomes yield 5-prime-biased transcripts, and you get vector everywhere with protein nowhere. SELEXON ADDS SEQUENCE TO A CASSETTE. So the question a gene-therapy-literate investor will ask in the Series A is: what does a selective exon cost us in kilobases, and what is our full-versus-empty and genome-integrity data on a SELEXON-containing vector at the intended dose. Avi and Hugh should hold that number and that assay before the question arrives; a splice element is small and the answer is probably fine, but "probably fine" is not a data-room answer. CALENDAR: Ultragenyx's decision on DTX401 (liver-directed AAV, glycogen storage disease type Ia, priority review, first therapy at the cause of the disease if cleared) is nominally due Sunday Aug 23, so realistically Friday Aug 21 or Monday Aug 24; our read is the temperature on systemically delivered AAV, nothing more. 1104HEALTH (docket pulled directly again this morning, HHSIG-2026-0034): STILL EIGHT COMMENTS, most recent posted Aug 14, and still NOT ONE from a trade association or major sponsor. TIMING CORRECTION: the Federal Register notice states 5:00 p.m. ET Monday Aug 24, while regulations.gov lists the comment period closing Aug 25 — work to Monday at five and do not spend the extra day. An eight-comment docket on a question sitting directly under 1104health's business model is the cheapest influence available to us this month, with three business days left. NO-CHANGE LINES: GPR52 schizophrenia NewCo (Lieber + Third Rock) — no GPR52 news in window, Nxera still has not named a buyer for NXE'149, clock unmoved. RNA translation-activator platform — nothing new since CAMP4's Q2; the Sept 28 analyst day and its EARLY PIPELINE remains the event. aSKY — no fresh KRAS catalyst; the G12D-selective competitive map is still the outstanding homework. Clasp — no update. Source in show notes: the SONIX Phase 1/2 protocol record for SKY-GJB2 (ClinicalTrials.gov). Link: https://clinicaltrials.gov/study/NCT07627971

Portfolio Watch — a competitor entered the clinic in Aletira's lead indication using Aletira's own architecture, and we found it a week late: Skylark Bio dosed the first child in SONIX with an engineered broad-transduction AAV capsid PLUS cell-specific expression control, delivering GJB2 to the cochlear supporting cells that natively make connexin 26 — the volume target in genetic deafness (200,000+ in US/EU) now has a clinical-stage first mover; plus Capricor reframes the FDA read (rigorous on evidence, flexible on path) ahead of Ultragenyx, the OIG docket sits nearly empty with 6 days left, and Eli Lilly funds a third mechanistic wave in schizophrenia

Tue, 18 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Tuesday, August 18, 2026. LEAD (a CATCH-UP — dated 2026-08-11, missed by our sweep last week; lands on flagship ALETIRA): Skylark Bio (Cambridge, MA; CEO Jodi Cook; scientific advisor Daniel Lee) came out of stealth and dosed the first patient in SONIX, an open-label multicenter Phase 1/2 of SKY-GJB2 in children 9 months to 7 years, with preliminary data guided for year-end 2026 and more in 2027. THE TARGET: GJB2 encodes connexin 26, a gap-junction protein; GJB2 mutations are the most common genetic cause of non-syndromic deafness — on the order of 200,000+ people across the US and Europe, no approved therapy addressing the cause. STAT called GJB2 the "holy grail" of hearing-loss targets. THE ARCHITECTURE IS WHY THIS MATTERS: a single unilateral intracochlear injection pairing an ENGINEERED AAV CAPSID DESIGNED FOR BROAD TRANSDUCTION with what Skylark calls CELL-SPECIFIC EXPRESSION CONTROL, to deliver functional GJB2 to the cochlear SUPPORTING CELLS that natively express connexin 26 and rebuild the gap-junction network. That is the SELEXON premise — get in broadly with the vector, then restrict where the payload is expressed — now in a child. THREE READS: (1) VALIDATION, uncomfortably precise: a clinical-stage company has bet a first-in-human program on exactly our division of labor, so the thesis that expression control is the differentiating layer gets clinical proof on someone else's dime; belongs in Geoff Lynn's Series A deck as a field-convergence slide, not surfaced by an investor across the table. (2) COMPETITIVE: the field has consolidated onto GJB2. Otoferlin is spoken for — Regeneron's Otarmeni took accelerated approval 2026-04-23 as the first gene therapy through the national priority voucher program (~50 US babies/year) — and Sensorion discontinued its own OTOF program in June to name its GJB2 candidate lead. So the small beachhead is taken and the large target now has a clinical-stage first mover plus a funded European fast-follower; if hereditary hearing loss is still Aletira's lead indication signal, that lane got materially more crowded. (3) TECHNICAL (Avi + Hugh Wells, this week): connexin 26 must go into supporting cells, NOT hair cells, at roughly the right level — gap junctions are dose-sensitive and overexpressing a connexin in the wrong cochlear cell is its own toxicity — the sharpest use case for splicing-based expression restriction in the ear. Skylark says "cell-specific expression control" but has NOT said how. If it is a cell-type-specific promoter, we have a real differentiation story (RNA-level fidelity, smaller cassette). If they have filed on splicing-based control, we must know BEFORE the raise. Pull the SONIX protocol and Skylark's patent estate. SECOND ITEM (fresh, 2026-08-14; reframes our standing Aletira read): Capricor stock +68% after its CEO said it will amend the deramiocel BLA (cardiosphere-derived cell therapy, Duchenne) with 24-month open-label extension data plus new analyses, NARROWING the indication to the primary endpoint it actually hit — upper-limb function; FDA indicated it will review the amendment and extend the PDUFA on receipt. Context: a July 30 adcomm voted 9-3 against effectiveness on a cardiac claim whose statistics FDA itself disputed. THE REFRAME (a correction to our own framing): this brief has cast the Ultragenyx decision as a read on whether the post-Elevidys SAFETY bar ratcheted up. True but half the picture. The Capricor sequence shows an agency RIGOROUS ON EVIDENCE and FLEXIBLE ON PATH — it holds you to what your statistics support and will not treat an adcomm vote as the end of the conversation if you narrow the claim to what you proved; consistent with the post-spring-turnover shift in rare-disease posture from adversarial toward partnership. WHY THAT IS A BETTER ARGUMENT FOR US: an agency that rewards narrowing rewards precision. A platform whose proposition is restricting EXPRESSION to the cells that need it is selling into a regulator currently rewarding sponsors for restricting CLAIMS to the patients they can prove — same instinct, one layer down. Stronger and more durable for Hemaka and Eddie than "the bar went up, so selectivity matters more." CALENDAR CORRECTION: Ultragenyx's Aug 23 is a SUNDAY and Capricor's Aug 22 is a SATURDAY. FDA does not act on weekends, so Ultragenyx realistically prints Friday Aug 21 or Monday Aug 24 — possibly the same day as our OIG filing. Staff a heavy Monday. 1104HEALTH (fresh, docket checked directly): the HHS OIG comment deadline has NOT moved (Monday Aug 24, end of day ET) and late comments are explicitly not accepted — but the docket has received a total of EIGHT comments and, as of Friday, NOT ONE is from a trade association or major sponsor (the only recent filing is a private individual in Maryland). With six days left the industry response is essentially unfiled, which turns this from compliance into LEVERAGE: on a nearly empty docket a specific, data-grounded comment from an operator who actually moves patients into trials carries disproportionate weight, and there is no basis for assuming the big trade groups carry the argument for us. File early enough to be read. Adjacent (2026-08-13, one day out of window): Sarah Cannon Research Institute + Merck collaboration to run Merck oncology studies at community sites — sponsor money moving into community-site trial delivery, which is 1104health's thesis. SCHIZOPHRENIA NEWCO (Lieber + Third Rock): NO GPR52 news at all this window — Nxera still has not named a buyer for NXE'149, so that clock has not moved. But a new mechanism arrived: on Aug 17 Leal Therapeutics closed a $30M Series A extension with ELI LILLY as a new investor and started a Phase 1b/2a in schizophrenia with an oral, brain-penetrant GLUTAMINASE (GLS1) inhibitor — a metabolic, glutamate-side mechanism that is neither dopaminergic nor muscarinic — initial data guided for year-end 2026. Two reads for Hemaka: Lilly funding a THIRD mechanistic wave says large-pharma appetite in schizophrenia is still wide open (good for partnering), but by the time our development candidate is declared later this year a competitor will have human data in a novel non-dopaminergic mechanism. That does not threaten our cognitive/negative-symptom differentiation, which is still the whole argument — but "novel mechanism" is getting less scarce, so the DC package should lead with three-domain data, not novelty. aSKY (landscape, 2026-08-17, the pairing is the lesson): AstraZeneca published two things the same day — SAFFRON, where Tagrisso plus a selective MET inhibitor (savolitinib) beat platinum doublet chemo on BOTH PFS and OS in EGFR-mutant NSCLC patients progressing on Tagrisso whose tumors carried high MET amplification/overexpression (338 patients, prospectively selected by IHC/FISH, 29 countries; ~34% of tumors take the MET resistance route after a 3rd-gen EGFR TKI) — and the DISCONTINUATION of eVOLVE-Lung02, its PD-1/CTLA-4 bispecific volrustomig in 1L NSCLC with PD-L1 under 50%, 895 patients, stopped because an independent committee judged it unlikely to hit either co-primary endpoint against pembrolizumab plus chemo. One company, one day, opposite outcomes, and the variable is PATIENT SELECTION: NSCLC value is created in prospectively selected molecular niches and destroyed in broad ones. So aSKY's undisclosed mechanism is now a TRIAL-DESIGN problem, not just a competitive-intelligence one — if that antibody has no companion selection hypothesis the team can name, it is being built for the side of the ledger that just lost. Krupnick and Eddie should answer on one slide: which patients, and by what assay. NO-CHANGE LINES: Clasp — no update; most recent is July's Clinical Cancer Research publication of CLSP-1025 preclinical data (framed as the first clinical-stage T-cell engager against an intracellular driver mutation); the p53 GUARDIAN-101 Phase 1 readout expected this year remains the event, watched via the Third Rock seat. RNA translation-activator platform — nothing new since yesterday's CAMP4 read; Sept 28 analyst day on the calendar. Gut-restricted GCPII/IBD — nothing in window; the IBD tape was quiet. Source in show notes: Skylark Bio SONIX first-patient-dosed announcement. Link: https://www.biospace.com/press-releases/skylark-bio-doses-first-patient-in-sonix-phase-1-2-trial-of-sky-gjb2-for-gjb2-related-pediatric-deafness

Portfolio Watch — a quiet direct-catalyst tape, but a real in-window read on our RNA lane: CAMP4's Q2 (Aug 13) shows the pace-setter funded through 2028 and cleared in Australia + Argentina, yet still NOT dosed — first-in-human is Q4 2026, correcting our own “already first-in-human” shorthand — and puts a Sept 28 analyst day on the board to map its early pipeline (the white-space call for our translation-activator platform); plus Ultragenyx's AAV gene-therapy FDA decision 6 days out and the HHS OIG trial-remuneration deadline 7 days out

Mon, 17 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for Monday, August 17, 2026 — another quiet direct-catalyst tape (no Blackbird company had a readout, financing, deal, or filing over the weekend), but for once a fresh, in-window item touching a program, plus a correction to our own briefing. LEAD (CAMP4 Therapeutics Q2 results, reported 2026-08-13; touches the RNA-based TRANSLATION-ACTIVATOR PLATFORM — CAMP4 is the pace-setter in the “turn-the-protein-up” / RNA-upregulation lane): three items worth attention. (1) Money settled — after the ~$50.1M second tranche closed early August, CAMP4 says cash runway now reaches END OF 2028; financing removed as a constraint for 2+ years. (2) Regulatory footprint widened on its lead asset CMP-002 (targets SYNGAP1-related neurodevelopmental disorder) — now cleared to start in BOTH Australia AND Argentina, with EU + UK filings submitted to open more sites. (3) Forward catalyst dated: an ANALYST DAY on Sept 28, 2026 covering trial design AND early discovery pipeline. CORRECTION (owed for precision): this brief has repeatedly described CAMP4 as “first-in-human and fully funded.” The “fully funded” half is right; the “first-in-human” half is NOT. CAMP4 has regulatory CLEARANCE to begin but has NOT dosed a patient — its own guidance is first-in-human INITIATES Q4 2026. So the leading independent RNA-upregulation company is on the clinic's doorstep, not through it — which gives our platform slightly more runway than the “already FIH” shorthand implied. SO-WHAT (Avi + platform leads): the Sept 28 analyst day is now the single most useful forward event on this beat — not the SYNGAP1 mechanics, but the EARLY DISCOVERY PIPELINE. CAMP4's platform “addresses >1,200 haploinsufficient disorders” is a framing number, not a pipeline; on Sept 28 they will likely NAME programs = a map of which diseases/tissues they plant their flag on next. That partially answers our 3-axis differentiation-wedge call for free (mechanism: genuine translation-step vs upstream regRNA/transcriptional; delivery: reach beyond CAMP4's intrathecal/CNS route, where liver/muscle/peripheral haploinsufficiencies are open white space; disease set: what we own that they don't). Action: put Sept 28 on the calendar; parse CAMP4's early pipeline the day it drops — by subtraction it shows where our white space actually is. STANDING BOARD: ALETIRA (flagship) — cleanest near-term read is 6 days out, the FDA's Aug 23 decision on Ultragenyx's liver-directed AAV gene therapy for glycogen storage disease type 1a; standing discipline holds (a climate read on the post-Elevidys gene-therapy safety bar, NOT a validation of SELEXON, which governs where a transgene is EXPRESSED, not where the capsid travels); fold the outcome + exact safety-label language into Geoff Lynn's Series-A narrative the day it prints. 1104HEALTH — HHS OIG comment deadline 7 days out (Aug 24); corrected framing holds (the RFI covers paying trial PARTICIPANTS/patients — travel, lodging, stipends — not paying physicians for coordination, which is the actual model); Rose Wang's team file by Aug 24 on the patient-remuneration question, play the physician-reimbursement question (roadmap-named, durable tailwind) as the longer structural track. GPR52 schizophrenia NewCo (Lieber + Third Rock) — no change; watch the deal tape (Nxera intends to out-license competing GPR52 agonist NXE'149 this half, “advanced discussions,” no buyer announced; a signed deal + buyer identity resets our differentiation clock). aSKY — no news; mutant-KRAS bar keeps rising, pressure concentrated in the G12D-selective subfield; cost of keeping our antibody's mechanism undisclosed rises monthly. Clasp Therapeutics (confirmed on our cap table) — no update; the p53 R175H (CLSP-1025) GUARDIAN-101 Phase 1 readout expected this year is the one to watch, via our Third Rock seat. Net: a quiet direct-catalyst day that isn't empty — CAMP4's quarter gives a real read on the RNA-upregulation pace-setter, corrects the “already first-in-human” shorthand, and puts a Sept 28 analyst day on the board; nearer in, Ultragenyx is 6 days out and the OIG deadline 7, both needing work this week. Source in show notes: CAMP4 Therapeutics Q2 2026 results (BioSpace, 2026-08-13). Link: https://www.biospace.com/press-releases/camp4-reports-second-quarter-2026-financial-results-and-corporate-highlights

Sourcing Radar — a real lead this week: Johns Hopkins (Kornberg + Calabresi) shows liver X receptor activation overcomes the inflammatory blockade of remyelination in multiple sclerosis, with human-cell validation, in-vivo myelin repair, and clean unencumbered IP — the build is a brain-penetrant, liver-sparing LXR-beta modulator; plus the deferred Elisseeff fibroid-senescence paper closed as a watch, not a build

Sun, 16 Aug 2026 12:00:00 +0000

The Blackbird Brief Sourcing Radar for Sunday, August 16, 2026 — after a thin prior week, a genuinely strong Johns Hopkins lead. LEAD (bioRxiv preprint, posted 2026-08-13; full text read): Michael Kornberg's lab at Johns Hopkins, with MS clinician-scientist Peter Calabresi on the author list, "Remodeling oligodendrocyte lipid metabolism via liver X receptors overcomes inflammatory blockade of remyelination." THE PROBLEM: every approved MS drug is immunomodulatory and cuts relapses/new lesions, but none repairs lost myelin — the failure that drives progressive disability. Oligodendrocyte precursor cells (the myelin-repair cells) remain abundant in lesions but are frozen by the inflamed environment, and some are flipped into an aberrant immune-like, antigen-presenting state that feeds inflammation. Prior remyelination bets (Biogen anti-LINGO/opicinumab, clemastine) could nudge differentiation under clean conditions but could not overcome the inflammatory brake — the whole game in a real patient. THE MECHANISM: interferon-gamma reprograms the precursor cell's lipid metabolism, flipping it from lipid synthesis to fatty-acid oxidation, depleting the fatty acids needed to build myelin and pushing the immune-like state; confirmed in mouse cells, human surgical specimens, and single-cell data from human MS lesions. THE FIX: a single upstream master switch — the liver X receptor (LXR) — controls this lipid program; pharmacologic LXR activation (tool agonists GW3965 and, structurally distinct, T0901317) rebalanced metabolism, suppressed the immune-like functions, and broke the interferon-gamma differentiation block in both mouse and human stem-cell-derived precursor cells. IN VIVO: in an adoptive-transfer/cuprizone model where inflammation specifically blocks repair, LXR activation increased mature oligodendrocytes and myelin repair — with dosing designed to start after immune infiltration so the benefit is shown to come from fixing repair cells, not additional immunosuppression. COMMERCIAL ANGLE: remyelination is the marquee validated whitespace in MS (no approved remyelinating therapy); fresh mechanism + human validation is exactly the seed-stage package a specialist neurology investor wants. IP IS CLEAN — the competing-interest statement discloses only unrelated senior-author consulting (Genentech, TG Therapeutics, OptumRx), NO equity, NO company, NO patent on this biology; funding is all society/foundation (National MS Society, Race to Erase MS, Johns Hopkins Catalyst Award, Gilbert Family Foundation, Maryland Stem Cell Research Fund) — an open, unencumbered Johns Hopkins tech-transfer license path (the Blackbird sweet spot). CANDID CATCH: this is a mechanism/target-validation paper, not a development candidate (tool compounds only), and LXR carries a known liability — broad LXR-alpha activation drives hepatic steatosis and hypertriglyceridemia, which stalled prior LXR programs. So the company is "build or in-license a brain-penetrant, liver-sparing LXR modulator against a newly validated rationale" — and the paper notes it is the LXR-beta isotype expressed in the repair cells, the natural handle for a selective, liver-sparing molecule. That selectivity problem is both the central risk and the moat. ACTIONS: Avi — is a beta-selective, CNS-penetrant LXR modulator tractable medchem, and who has chemical matter to in-license around; Eddie — open the Johns Hopkins tech-transfer field-of-use conversation now (provisional-filing moment; Kornberg + Calabresi is founding-team caliber). SECONDARY (closing last week's deferred lead): Jennifer Elisseeff's Johns Hopkins uterine-fibroid senescence atlas has now rendered and was read in full — a first single-cell map of senescent cells driving fibroid matrix stiffening and abnormal vasculature via a mechanosensing program, with a translational hook (collagenase from a Johns Hopkins fibroid trial reduced both matrix and senescent burden), in a huge underserved market (fibroids affect the majority of women; leading cause of hysterectomy; little non-surgical). But it is a WATCH, not a build: the IP is thin/derivative (descriptive atlas; the one therapy, collagenase, is already clinical and commoditized; the novel mechanosensing driver is a hypothesis), and the senior author is a serial senescence founder with equity in two senescence/soft-tissue companies and a consulting role in a third — a heavy encumbrance flag that any fibroid-senescence IP likely reads onto. Keep her on the watch list. WHAT ELSE CROSSED (not leads): a Johns Hopkins 20-marker serial-IHC pathology tool (useful but undefendable — off-the-shelf reagents, crowded field: Akoya, Lunaphore; services angle only); medRxiv was all public-health/informatics with no asset; Lieber Institute and Kennedy Krieger were silent (no corresponding-author preprints in-window). Sources in show notes: the Kornberg LXR remyelination bioRxiv preprint. Paper link: https://www.biorxiv.org/content/10.64898/2026.08.07.743529v1

Portfolio Watch — a fifteenth straight quiet catalyst day, so a Sunday week-ahead: the two hardest dates on our board have arrived inside eight days and each needs work this week — Ultragenyx's AAV gene-therapy FDA decision Aug 23 (a post-Elevidys safety-climate read for Aletira and a timing input for Geoff Lynn's raise) and the HHS OIG trial-remuneration comment deadline Aug 24 (a filing from 1104health's team on the patient-payment question)

Sun, 16 Aug 2026 17:00:00 +0000

The Blackbird Brief Portfolio Watch for Sunday, August 16, 2026 — the fifteenth straight quiet direct-catalyst day: no Blackbird company had a readout, financing, deal, or filing this weekend, and the only fresh in-window item touching our geography was a regional roundup with nothing on a Blackbird program. Rather than manufacture a catalyst, a Sunday week-ahead anchored on the two hard dates now inside the next eight days, each needing action this week. LEAD (calendar/climate read, lands on flagship ALETIRA): the FDA decision on Ultragenyx's liver-directed AAV gene therapy for glycogen storage disease type 1a is due Saturday Aug 23 — 7 days out. Significance: it is the first FDA verdict on a systemic AAV gene therapy since the field's safety picture darkened over the summer (Elevidys deaths + boxed warning, the China pediatric gene-editing deaths, Intellia's liver-tox tied to a patient HLA variant) — a live test of how far the agency's risk tolerance has actually moved. Outcome-conditional read: a clean approval says the bar is still efficacy-plus-manageable-safety (field keeps moving); an approval with a tight label / boxed warning / heavy postmarketing, or a delay/CRL, says the precision-and-safety bar has ratcheted up = the tailwind under Aletira's SELEXON story. STANDING CAVEAT (kept, because a sharp investor pulls the seam): SELEXON governs where a transgene is EXPRESSED (which cell type reads it out), NOT capsid biodistribution (where the virus travels), which is the axis most of these safety events turned on — so this is a CLIMATE read on where Geoff Lynn is raising into, NOT proof SELEXON would have prevented them. Action (Hemaka + Eddie): read the Aug 23 outcome and the exact safety label language as a TIMING input for sequencing Geoff's raise. SECOND DATE (needs a filing from us), lands on 1104HEALTH: the HHS OIG comment deadline on anti-kickback safe harbors in clinical trials is Monday Aug 24 — 8 days out. Corrected framing held: this RFI is about paying trial PARTICIPANTS/patients (travel, lodging, stipends, completion incentives), NOT paying physicians for the coordination work that is 1104health's core model — two different payment questions on two clocks. The patient-remuneration question closes Aug 24; the physician-reimbursement question (our actual thesis) runs on a slower structural track through the agency and is now a durable tailwind (the government's own trial-reform roadmap names that reimbursement gap explicitly). Action (Rose Wang's team): file by Aug 24 on the patient-remuneration question (a community-oncology marketplace has a real-world view worth putting on the record); play the physician-coordination question as the longer game. STANDING BOARD (all no-change, one line each): GPR52 schizophrenia NewCo (Lieber + Third Rock) — watch the deal tape; Nxera intends to out-license its competing GPR52 agonist NXE'149 this half and is in "advanced discussions," but no buyer announced — a signed deal (and the buyer's identity) resets our differentiation clock. aSKY — no news; the mutant-KRAS bar keeps rising (G12D-selective field is the pressure point; cost of keeping our antibody's mechanism undisclosed rises monthly). Clasp Therapeutics (confirmed on our cap table) — no update; the p53 R175H (CLSP-1025) GUARDIAN-101 Phase 1 readout expected this year is the one to watch, via our Third Rock seat. RNA translation-activator platform — no change (CAMP4 first-in-human and fully funded; the 3-axis differentiation wedge stays this cycle's live task). Gut-restricted GCPII/IBD (Slusher) — no change (TL1A wave keeps crowding the anti-inflammatory bucket; differentiate on the enteric-pain axis). Net: quiet ends next week — two hard dates back-to-back (Ultragenyx Aug 23, OIG Aug 24), both needing work THIS week. Sources in show notes: Ultragenyx investor relations (DTX401 PDUFA Aug 23). Link: https://ir.ultragenyx.com/

Portfolio Watch — another quiet catalyst tape, so I closed yesterday's ownership question: the confirmed Blackbird KRAS check runs through Clasp Therapeutics, a Johns Hopkins-born T-cell-engager spinout we backed at launch — which means Blackbird now holds three KRAS modalities (antibody via aSKY, vaccine via Adventris, TCR-mimic bispecific via Clasp) across the whole disease spectrum, with a Clasp p53 Phase 1 readout due this year and a Third Rock seat to watch it through

Sat, 15 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 15, 2026 — another quiet direct-catalyst day (no Blackbird company had its own readout, financing, deal, or filing overnight). Rather than a countdown, this episode closes the ownership question left open in the 08-14 Adventris episode ("is it ours?") and, in doing so, surfaces a confirmed Blackbird portfolio company that was missing from the brief's own map. LEAD (Clasp Therapeutics — a confirmed Blackbird BioVentures portfolio company not previously tracked here): Clasp is a Johns Hopkins spinout built on the work of Bert Vogelstein and Drew Pardoll (Bloomberg~Kimmel immunotherapy institute — the same Hopkins bench Adventris traces to), which launched publicly in March 2024 with a $150M Series A led by Catalio, Third Rock Ventures, and Novo Holdings; Blackbird BioVentures is a NAMED participant in that financing (confirmed via the company's launch materials and JHTV's announcement). Maryland nexus: offices in Cambridge, MA and Rockville, MD. NOT yet pinned down: Blackbird's check size / ownership % / follow-on (Eddie + Emily to pull from our records). THE SCIENCE: Clasp's pHLAre platform builds TCR-mimic bispecific T-cell engagers ("precise HLA redirecting engagers") that, unlike surface-antigen TCEs, target the tumor-specific peptide fragments of mutant driver proteins presented on HLA — one arm clamps the mutant-peptide/HLA complex, the other grabs CD3 on a T cell, killing only mutation-bearing cells (off-the-shelf, antibody-like, absolute-specificity pitch). PIPELINE / two KRAS-board hooks: (1) lead program CLSP-1025 targets the p53 R175H mutation on HLA-A*02:01, in the first-in-human GUARDIAN-101 Phase 1 in advanced solid tumors (first patient dosed ~April 2025) — a monotherapy readout expected THIS YEAR is a genuine near-term catalyst and the first in-patient test of the platform; (2) CLSP-5282, a KRAS G12V T-cell engager on HLA-A*03:01, unveiled at AACR 2026 (April) and now in its own Phase 1 (SENTINEL-101) — preclinically it kills HARDER against KRAS-inhibitor-resistant tumors and is enhanced by combination with small-molecule KRAS inhibitors, i.e. positioned to layer on top of / after the daraxonrasib-class drugs, not compete head-to-head. SO-WHAT (the corrected KRAS map): Blackbird now has THREE distinct modalities against mutant KRAS spanning the full disease timeline — aSKY (UMB, antibody vs an undisclosed KRAS frame; our owned program), Adventris (BioHub, synthetic long-peptide vaccine; adjuvant/interception), and Clasp (TCR-mimic bispecific TCE; established, treatment-refractory disease). The field I've called "hardening" around Revolution Medicines' small molecules and degraders now maps as antibody / small molecule / degrader / vaccine / TCE — and Blackbird has a stake in three of the five, at least one (Clasp) confirmed-funded. CONNECTIVE TISSUE: Third Rock co-leads Clasp AND is our venture partner on the GPR52 schizophrenia NewCo (Lieber Institute) — a deliberate channel for read-through on the Clasp Phase 1 and Third Rock's KRAS-immunotherapy thinking. ACTIONS: (a) Eddie + Emily — pull our actual Clasp position (check size, ownership %, follow-on) and resolve the Adventris ownership question against the same records; (b) aSKY competitive map — add a T-cell-engager column (Clasp) alongside the vaccine/interception column added 08-14; the Clasp p53 Phase 1 readout (this year) is the next real datapoint and it's ours to watch via Third Rock. BOARD (all no-change): Aletira — FDA's Aug 23 decision on Ultragenyx's liver-directed AAV gene therapy for glycogen storage disease type Ia is 8 days out; a climate read on the post-Elevidys safety bar, NOT SELEXON validation (SELEXON governs expression selectivity, not capsid biodistribution). 1104health — the OIG comment deadline (Aug 24, 9 days) is about paying trial PARTICIPANTS/patients, not physicians for coordination; comment on the patient-remuneration question, play the physician-reimbursement question as the longer structural track. GPR52 schizophrenia NewCo (Lieber + Third Rock) — watch the deal tape (Nxera out-licensing competing NXE'149 this half). RNA translation-activator platform — no change (CAMP4 first-in-human and funded; wedge urgent). Gut-restricted GCPII/IBD — no change (TL1A wave crowding; differentiate on the enteric-pain axis). Sources in show notes: Johns Hopkins Technology Ventures Clasp launch announcement; Clasp $150M Series A launch release (BioSpace/BusinessWire, naming Blackbird BioVentures); Clasp CLSP-5282 AACR 2026 release; GUARDIAN-101 (CLSP-1025) Phase 1. Link: https://ventures.jhu.edu/news/clasp-therapeutics-launches-with-150-million-investment-in-pioneering-precision-immuno-oncology/

Portfolio Watch — a quiet catalyst tape, so I closed a gap on our own board: Adventris, a Johns Hopkins-licensed KRAS-vaccine company resident in the Blackbird BioHub, just put three phase 1 readouts on the table — including a differentiated cancer-interception result the rest of the KRAS field can't reach; the aSKY competitive map now needs a vaccine column

Fri, 14 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 14, 2026 — another quiet direct-catalyst day (no Blackbird company had its own readout, financing, deal, or filing overnight). Rather than a fifth straight countdown to the two August dates, this episode closes a gap in the brief's own field of view, surfaced while doing the aSKY KRAS-competitive-map homework flagged Tuesday. LEAD (Adventris Pharmaceuticals — a Blackbird BioHub company we haven't been tracking): Adventris is a Baltimore company built on a Johns Hopkins platform, co-founded by Elizabeth Jaffee and Neeha Zaidi (Bloomberg Kimmel Institute for Cancer Immunotherapy), which licensed the technology from Johns Hopkins. The asset, mKRAS-VAX, is an off-the-shelf synthetic long-peptide vaccine presenting the six most common KRAS mutations (the G12 variants plus G13D) to train T cells against the mutation itself — not a small molecule, not an antibody, a vaccine. Confirmed: Adventris is a resident at the Blackbird BioHub and was featured at the spring showcase, and it's built on Hopkins-licensed IP; NOT independently confirmed: a Blackbird BioVentures investment — so treat it as squarely in Blackbird's orbit / on its real estate, and confirm the exact ownership line. THE DATA — three phase 1 readouts staking out three points on the disease timeline (per PubMed): (1) resected pancreatic cancer, vaccine + ipilimumab/nivolumab, 11/12 patients mounted a significant mutant-KRAS T-cell response, grade 1-2 AEs (Nat Commun 2026, DOI 10.1038/s41467-026-68324-4); (2) metastatic MMRp/MSS colorectal cancer (checkpoint-refractory), vaccine + nivolumab/ipilimumab, T-cell responses in 75% of evaluable patients ex vivo and 100% after expansion (Nat Commun 2026, DOI 10.1038/s41467-026-74711-8); (3) THE ONE TO CIRCLE — cancer interception/prevention, published in Cancer Discovery mid-July: 20 individuals with hereditary PDAC predisposition and a radiographic pancreatic lesion (no cancer yet) given the vaccine ALONE (no checkpoint), 18/20 (90%) generated a significant mutant-KRAS T-cell response, clones persisted up to 2 years, cyst reduction/resolution in ~38% vs ~7% unvaccinated, and no participant developed PDAC over ~16.5 months median follow-up (DOI 10.1158/2159-8290.CD-25-2245). Caveats: 20 patients, single-arm, immune-response endpoint, prevention is a long/expensive endpoint to prove. SO-WHAT (ties to aSKY): the KRAS landscape I've described as hardening — Revolution Medicines' daraxonrasib (survival doubled in previously-treated metastatic PDAC, now under FDA review), G12D-selective zoldonrasib (Breakthrough designation, three Phase 3 starts), Incyte's G12D inhibitor, a G12D degrader — is entirely aimed at established, metastatic disease, and aSKY's undisclosed KRAS-frame antibody must define against that armada. But the map stopped at treatment. The vaccine modality (Adventris) plays a different board: adjuvant, minimal-residual-disease, and outright interception in high-risk people — settings a metastatic-disease inhibitor was never designed for. Corrected read: Blackbird has KRAS exposure on two modalities across three-to-four disease settings, one of them literally down the hall. ACTION (Eddie + team): add a vaccine-and-interception column to the aSKY competitive map, and confirm whether Adventris is ours, partly ours, or a neighbor whose success validates the same target thesis. BOARD (all no-change): Aletira — FDA's Aug 23 decision on Ultragenyx's liver-directed AAV gene therapy for glycogen storage disease type Ia is 9 days out; a climate read on the post-Elevidys safety bar, NOT SELEXON validation (SELEXON governs expression selectivity, not capsid biodistribution); fold the terms into the Series A narrative. 1104health — the OIG comment deadline (Aug 24, 10 days) is about paying trial PARTICIPANTS/patients, not physicians for coordination; comment on the patient-remuneration question, play the physician-reimbursement question as the longer structural track. GPR52 schizophrenia NewCo (Lieber + Third Rock) — watch the deal tape (Nxera out-licensing competing NXE'149 this half). RNA translation-activator platform — no change (CAMP4 first-in-human and funded; wedge urgent). Gut-restricted GCPII/IBD — no change (TL1A wave crowding; differentiate on the enteric-pain axis). Sources in show notes: three JHU/Adventris mKRAS-VAX phase 1 papers (Cancer Discovery interception study, two Nature Communications treatment studies, via PubMed); Blackbird BioHub showcase coverage (BioBuzz). Link: https://doi.org/10.1158/2159-8290.CD-25-2245

Portfolio Watch — a correction, not a catalyst: the Aug 24 OIG comment deadline we've flagged for 1104health is about paying PATIENTS, not paying physicians for coordination — so it doesn't de-risk the core model the way I said; but Operation TrialBlazer just put the physician-reimbursement gap (1104health's whole reason to exist) in writing

Thu, 13 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 13, 2026 — another quiet direct-catalyst day (no Blackbird company had its own readout, financing, deal, or filing), but today is a correction, not a countdown rerun, and it changes an action item due in 11 days. LEAD (1104health — a correction to this brief's own standing framing): for weeks I've flagged the HHS Office of Inspector General's August 24 comment deadline as load-bearing for 1104health, framed as the anti-kickback question of whether pharma-funded subsidies to community oncologists for trial-coordination work sit cleanly inside the safe harbors, and told Rose Wang's team to file a letter as insurance for exactly that. On a closer read — and the legal analysis across several firms (Hall Render, Sidley, Holland & Knight) is unanimous — the OIG Request for Information is about remuneration to trial PARTICIPANTS (the patients): travel, lodging, childcare, meal costs, stipends for time, completion incentives. It is explicitly NOT about paying physicians or sites for coordination. So the deadline is real but governs the patient side, not the physician-subsidy mechanism that is the core of 1104health's revenue model — I conflated the two. The corrected map: two separate payment questions inside Operation TrialBlazer on two different clocks — (1) patient remuneration (the Aug 24 RFI), and (2) the physician/site reimbursement gap, which is 1104health's thesis. The genuinely good news: the TrialBlazer roadmap NAMES the physician-reimbursement gap explicitly ("participating in a clinical trial creates administrative burden, workflow disruption and lacks a reimbursement structure that reflects that investment," especially in community care) — 1104health's thesis stated back by HHS = a durable tailwind. But that physician-side question does NOT get resolved by Aug 24; it's a slower structural track running through Medicare (CMS) and the OIG together, on a quarters-not-weeks timeline. Revised action, two parts: (a) 1104health should still comment by Aug 24 but on the patient-remuneration question specifically (patient stipends/travel/completion incentives shape enrollment + retention economics inside a community-oncology marketplace, and Rose Wang's team has a real-world view worth putting on the record); (b) do NOT treat Aug 24 as the day the physician-subsidy compliance question closes — it's alive, now explicitly validated as a barrier the government wants solved, and the right posture is to engage the longer reimbursement-structure track. Net: the thesis tailwind got stronger this week; the near-term regulatory de-risking got weaker — the opposite of what my earlier framing implied. BOARD (mostly no-change): Aletira — the FDA's Aug 23 decision on Ultragenyx's liver-directed AAV gene therapy for glycogen storage disease type Ia is still the top read, 10 days out; a climate read for the SELEXON Series-A narrative (NOT validation — SELEXON governs expression selectivity, not capsid biodistribution), with the terms of any liver-safety conditions the cleanest near-term signal of the post-Elevidys safety bar. aSKY — no change to the core read; one adjacent datapoint: Eli Lilly's olomorasib got FDA Breakthrough Therapy designation in previously-treated KRAS G12C pancreatic cancer (a different mutation than the G12D subfield aSKY's frame points at, so not a direct comparator, but one more marker the mutation-selective RAS field is de-risking fast); action stands — refresh the competitive map against the G12D-selective set. GPR52 schizophrenia NewCo (Lieber + Third Rock) — no news; watch the deal tape (Nxera intends to out-license competing NXE'149 this half). RNA translation-activator platform — no change (CAMP4 first-in-human + funded; wedge task urgent). Gut-restricted GCPII/IBD — no change (TL1A wave crowding; differentiate on the enteric-pain axis). Sources in show notes: OIG RFI on anti-kickback safe harbors for clinical-trial participant remuneration (comments due Aug 24, 2026); HHS Operation TrialBlazer roadmap; Ultragenyx DTX401 FDA priority review (PDUFA Aug 23, 2026). Link: https://hallrender.com/2026/06/26/oig-issues-rfi-on-anti-kickback-statute-safe-harbors-for-clinical-trial-participant-remuneration-comments-due-august-24-2026/

Portfolio Watch — an honest quiet tape, but two hard dates now inside two weeks: Ultragenyx's gene-therapy decision (Aug 23), the cleanest live read on the post-Elevidys safety bar Aletira is raising into, and the OIG anti-kickback comment deadline (Aug 24) that needs a letter from us now; plus a sharpened aSKY read — the KRAS G12D-selective field, not just pan-RAS, is the bar

Wed, 12 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 12, 2026 — another quiet direct-catalyst day (no Blackbird company had its own readout, financing, deal, or filing), and a full sweep of every lane (oncology, schizophrenia, gene therapy, IBD, the RNA platforms, and the Baltimore tape) turned up nothing genuinely new in the last few days. No manufactured lead; instead a countdown and a board check. LEAD (landscape read on the flagship, Aletira — a climate read, NOT a portfolio catalyst): the FDA's decision date on Ultragenyx's DTX401 AAV gene therapy for glycogen storage disease type Ia is August 23, now 11 days out. It is the cleanest, closest test of where the FDA's gene-therapy safety bar sits after the Sarepta Elevidys collapse (patient deaths from acute liver failure, boxed warning, eligibility pulled, platform designation revoked) — a liver-directed AAV therapy going up for approval into a chastened climate, so the outcome AND its terms (label restrictions, monitoring/risk-management conditions, any liver-safety warning) are a live datapoint on how high the bar now sits for the whole modality. So-what for Aletira (same discipline as always): a climate read, not validation — SELEXON governs where a transgene is EXPRESSED (cell-type-selective via alternative RNA splicing), not capsid biodistribution, and much of Elevidys liver-tox was capsid-dose/biodistribution-driven; so an eventful decision would not prove SELEXON would have prevented it. What it sets is the temperature of the room Geoff Lynn raises the Series A into — heavy safety conditions = precision is the price of admission (the climate a selectivity platform wants), a clean approval = a subtler read. Action: Hemaka + Eddie fold the decision and its precise terms into the raise narrative as a real-time market signal. SECOND date (needs a document from us): the HHS Office of Inspector General comment window on the anti-kickback safe harbors — the Operation TrialBlazer piece bearing on 1104health — closes August 24, 12 days out. 1104health puts trial enrollment in community oncologists' hands at the point of care; whether pharma-funded physician subsidies sit cleanly inside the safe harbors is the one regulatory question that touches the heart of the model, and the comment deadline is the moment to get our position on the record. Rose Wang's team owns the draft; the drafting window is now. BOARD (one line each, with one genuinely sharpened read): GPR52 schizophrenia NewCo (Lieber + Third Rock) — no news; standing watch is the deal tape (Nxera's stated intent to sign a major pharma partner for competing NXE'149 this half is the event that resets our clock). aSKY — SHARPENED: the competitive bar isn't just pan-RAS daraxonrasib anymore, it's the KRAS G12D-SELECTIVE subfield aSKY's own KRAS frame points at, and that subfield hardened — Revolution Medicines' G12D-selective zoldonrasib now carries FDA Breakthrough Therapy designation in G12D lung cancer and moves into three Phase 3 trials this year (two first-line pancreatic, including a KRAS-G12D-specific study, one first-line lung), plus a G12D-selective inhibitor from Incyte and a G12D-targeted degrader now in the clinic; Eddie + aSKY refresh the competitive map against the G12D-selective set specifically, and the cost of keeping the mechanism undisclosed keeps rising. RNA translation-activator platform — no change (CAMP4 first-in-human and funded; differentiated-wedge task urgent). Gut-restricted GCPII/IBD — no change (TL1A wave keeps crowding, with best-in-class Phase 2b data behind the Sanofi/Teva antibody and Merck's Phase 3 win; differentiate on the enteric-pain axis). Sources in show notes: Ultragenyx DTX401 FDA acceptance and priority review (PDUFA Aug 23, 2026); Revolution Medicines zoldonrasib FDA Breakthrough Therapy designation and Phase 3 program. Link: https://www.biospace.com/press-releases/ultragenyx-announces-u-s-fda-acceptance-and-priority-review-of-the-biologics-license-application-bla-for-dtx401-aav-gene-therapy-for-glycogen-storage-disease-type-ia-gsdia

Portfolio Watch — Nxera's GPR52 rival moves from shelved to a late-stage deal process (out-license intent this half), putting our schizophrenia NewCo's differentiation clock on a live timer

Tue, 11 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 11, 2026 — a sixteenth straight quiet direct-catalyst day (no Blackbird company had its own readout, financing, deal, or filing), but the competitive tape moved on the same program that led yesterday: the GPR52 schizophrenia NewCo (Lieber Institute + Third Rock). Lead (competitive-intelligence, not a portfolio catalyst): on Friday, August 7, Nxera Pharma (formerly Sosei Heptares) reported Q2 results and held its earnings call, and management gave the most concrete update in months on NXE'149 — the only other Phase-2-ready GPR52 agonist in industry and the direct same-mechanism rival to our program. Trajectory is the story: Boehringer Ingelheim optioned NXE'149, was expected to license it in late 2025, then walked away in December 2025; rights reverted to Nxera and the asset has been quietly for sale (and unsold) since. Friday it changed. Management said on the record they are in "advanced discussions with several parties" to partner the program and expect to execute at least one major out-license in the second half of 2026; CEO stated it is "not our intention to conduct global phase II studies ourselves" — they want a large, fully-funded pharma partner to carry it into Phase 2/3. So this moved from a passive shopping process (dragging since December) to an active, late-stage deal process with a stated close window measured in months. So-what for our NewCo (this is the calendar risk we wrote down): if a buyer picks up NXE'149 this half, our Lieber NewCo enters its development-candidate declaration (expected later 2026) 12-24 months behind a direct, same-mechanism competitor now backed by a large pharma balance sheet — the thing that most cleanly caps the value of being early on GPR52. Two sharpeners: (1) our "best-in-class GPR52 chemistry / first-differentiated-Phase-2" framing only holds while nobody else is in Phase 2, so its shelf life is now this half, not open-ended; (2) the pressure is bilateral — on the same call Nxera flagged that muscarinic momentum (MapLight's Phase 2 readout, yesterday's lead) has strengthened the competitive picture, so the field is consolidating from two directions at once (a same-mechanism buyer plus a validated adjacent class). Strategic wrinkle: Nxera positions GPR52 for schizophrenia AND psychosis in Alzheimer's disease (very large, no approved drug) — a broad-indication buyer validates the target but raises the buyer's incentive to move fast, compressing our window further. Actions (Hemaka + Third Rock, ahead of the DC declaration): lock the best-in-class chemistry story and first-differentiated-Phase-2 framing now (depreciating in real time); have the hard conversation on whether to compress DC timing to get ahead of a Nxera signing versus holding for the cleaner molecule; watch the deal tape — a Nxera partner announcement (and who it is) resets our clock precisely and tells us which indications/geographies get contested. Board (one line each): Aletira — no fresh news; Ultragenyx AAV decision Aug 23 (now 12 days out) is the live gene-therapy safety-bar read, the climate a selectivity platform wants to raise into; RNA translation-activator — no change (CAMP4 first-in-human and fully funded, wedge task urgent), with a landscape caution that the RNA-therapeutics tape isn't uniformly hot (Alnylam shed roughly $12B in market value last week on core-business questions); 1104health — HHS OIG anti-kickback safe-harbor comment deadline Aug 24 (13 days out) needs a document from Rose Wang's team now; aSKY — nothing new since Revolution Medicines' quarter, mutant-RAS reference-setter keeps de-risking the small-molecule lane and the bar keeps rising; gut-restricted GCPII/IBD — TL1A crowding, differentiate on the enteric-pain axis. Sources: Nxera Pharma Q2 2026 results and earnings call (GlobeNewswire / Investing.com transcript, August 7, 2026). Link: https://www.globenewswire.com/news-release/2026/08/07/3340927/0/en/nxera-pharma-operational-highlights-and-consolidated-results-for-the-second-quarter-2026.html

Portfolio Watch — MapLight's Phase 2 muscarinic win plus Cobenfy's soft launch raise the tolerability-and-cognition bar for our GPR52 schizophrenia NewCo

Mon, 10 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 10, 2026 — a fifteenth straight quiet direct-catalyst day (no Blackbird company had its own readout, financing, deal, or filing over the weekend), but with a genuinely fresh, substantive competitive datapoint that lands on the schizophrenia NewCo (Lieber Institute + Third Rock). Lead (landscape/competitive, not a portfolio catalyst): today's trade coverage centers on MapLight Therapeutics (Series D of $372.5M last summer, co-led by Forbion and Goldman Sachs). Candid on recency — MapLight's Phase 2 ZEPHYR readout itself dropped July 27; what's new today is the analysis pairing it with hard Cobenfy commercial numbers, and this brief has never covered the MapLight data. The data: ML-007C-MA is an M1/M4 muscarinic agonist co-formulated with an anticholinergic (same class/trick as Bristol Myers' Cobenfy, xanomeline-trospium). In ZEPHYR (307 inpatients, acute psychosis), the twice-daily dose hit the primary endpoint — 4.5-point PANSS separation vs placebo at week 5, effect size ~0.37, ~6.0-point separation in completers; the once-daily arm MISSED. Two standouts: clean tolerability (only 2% GI discontinuations, no metabolic signal, no EPS, no urinary retention) and a cognition signal (effect size ~0.5 in baseline-impaired patients, decorrelated from PANSS improvement). Caveats: 4.5 points is modest (below Cobenfy's registration trials) and the QD arm failed. The reality check next to it: Cobenfy — first-in-class non-dopaminergic antipsychotic, first new mechanism in 35 years — booked only $155M in 2025 vs forecasts north of $310M, dragged by GI tolerability and access. So-what for the GPR52 NewCo (bar-raising, not a threat to the thesis): MapLight is NOT a same-mechanism rival like Nxera (muscarinic vs our orphan-GPCR agonist), but it puts a funded, clinically-validated competitor on the two axes our value case rests on. (1) Tolerability — a clean mechanism doesn't sell itself; MapLight showed you can engineer a muscarinic to clean tolerability, so "better tolerated than old atypicals" is now table stakes, and our clean-GPCR tolerability story has to be at least as good. (2) Cognition — the reason GPR52 justifies a NewCo is reaching all three symptom domains; MapLight just planted a flag on independent cognition, so "we also reach cognition" is no longer clean white space. Defensible story can't be "non-dopaminergic + hits cognition"; it has to be materially bigger/broader efficacy than a muscarinic's modest 4.5-point signal, or a cognition/negative-symptom mechanism muscarinics can't reach. Actions (Hemaka + Third Rock, ahead of the later-2026 DC declaration): target product profile must answer MapLight directly on tolerability and the breadth of the cognition claim; fold Cobenfy's $155M reality into the value case (novelty/first is not the moat — design commercial/access in now); Nxera's GPR52 program still unsold, so no same-mechanism race yet, but the window to define best-in-class is not indefinite. Board (one line each): Aletira — no fresh news; Ultragenyx AAV decision Aug 23 is the live safety-bar read, sharpened by Friday's Intellia HLA disclosure toward precision; RNA translation-activator — CAMP4 FIH + fully funded, differentiated-wedge task live/urgent; 1104health — HHS OIG anti-kickback safe-harbor comment deadline Aug 24 needs a document from Rose Wang's team now; aSKY — nothing new since RevMed's quarter, mutant-RAS bar keeps rising; gut-restricted GCPII/IBD — TL1A crowding, differentiate on the enteric-pain axis. Sources: BioSpace (Aug 10, 2026) on Cobenfy traction + MapLight; MapLight ZEPHYR topline (July 27, 2026).

Sourcing Radar — a thin week: a novel heme-transporter target (SLC49A3) out of UMB, encumbered by the PI's own company; plus a JHU/UMB fibroid-senescence atlas held for next week

Sun, 09 Aug 2026 13:00:00 +0000

The Blackbird Brief Sourcing Radar for August 9, 2026 — a thin partner-institution week. The corresponding-author sweep across Johns Hopkins, UMB, and the Lieber Institute for the past seven days turned up no de-risked asset (no molecule, no in-vivo efficacy, nothing licensable-and-buildable now). Lead (full-text read, from Iqbal Hamza's lab, University of Maryland School of Medicine, bioRxiv v1 dated 2026-08-03): a genetic screen in C. elegans shows loss of the intestinal heme exporter MRP-5 causes lethal endolysosomal heme trapping; the bypass is a previously uncharacterized SLC49A-family importer-to-exporter handoff, and they establish the human transporter SLC49A3 as a conserved heme exporter — knockdown causes heme overload, premature hemoglobinization, and death in zebrafish and human erythroid precursors, and it has a heme-binding pocket distinct from the known FLVCR/HRG1 nodes. The paper explicitly frames the circuit as targetable "to correct heme maldistribution in hematologic and mitochondrial disease" — a genuinely novel druggable transporter class relevant to anemias/erythropoiesis and mitochondrial disease. Candid catch (why it's watch, not build): it is early target-ID (worm to zebrafish to cultured human cells), no molecule and no disease-linked in-vivo efficacy — and the competing-interest statement says the corresponding author "is the President and Founder of Rakta Therapeutics Inc., a company involved in the development of heme transporter-related diagnostics." Rakta is described as diagnostics, so a therapeutic on SLC49A3 may be open ground, but you cannot evaluate a clean Blackbird build without first understanding what Rakta controls, what UM Ventures has filed, and the inventor's availability/allegiance — a freedom-to-operate question you answer first. Actions: Eddie — call UM Ventures on the therapeutic-target filing and whether it sits inside or outside Rakta; Avi — is a transporter like this tractable as a small-molecule/biologic target. Second item, flagged for next week (NOT lead-eligible today): Jennifer Elisseeff's lab (JHU; proven serial founder) posted a cellular-senescence atlas of uterine fibroids (bioRxiv v1 2026-08-07) mapping the senescent cells that drive matrix stiffening and abnormal vasculature and naming them as targets for non-surgical, senescence-directed therapy — a large unmet-need, anti-fibrotic/senotherapeutic angle with live Hopkins-Maryland (MPower/Pivot) funding. But the v1 body is not yet rendered (abstract only, transient), so per the full-text eligibility gate it cannot lead this week — revisit on the next re-pull; even fully read it is a would-take-the-meeting on market + founder, not an asset. What else crossed and why it's not a lead: mostly JHU methods/tools (protein-DL benchmarks, spike-sorting, spatial-transcriptomics models); a UMB nanoparticle hepatitis-C vaccine (generic scaffold, weak commercial pull given curative antivirals); medRxiv entirely clinical-epi/devices. Paper link: https://www.biorxiv.org/content/10.64898/2026.07.31.742071v1

Portfolio Watch — Intellia ties an HLA immune-gene variant to the worst liver tox from its in-vivo CRISPR therapy; a fresh, precision-flavored read for Aletira

Sun, 09 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 9, 2026 (the Sunday double, second half) — a fourteenth straight quiet direct-catalyst day (no readout, financing, deal, or filing landed squarely on a Blackbird company this week), but with one genuinely fresh, in-window item that lands on the flagship Aletira. Lead (landscape/watch, not a portfolio catalyst): in its August 6 Q2 update, Intellia Therapeutics disclosed a genetic explanation for the liver toxicity that dogged nex-Z, its systemically (lipid-nanoparticle) delivered in-vivo CRISPR therapy for transthyretin amyloidosis (ATTR) — an analysis of 600-plus patient samples found the worst transaminase elevations concentrate in carriers of one specific HLA immune-gene variant (HLA-C*05:01); the CEO's words: "each of the five highest elevations observed following dosing occurred in patients carrying this allele," though the strong majority of carriers were unaffected. Context: two Phase 3 trials paused October 2025, a patient death November 2025, hold lifted March 2026 with enhanced monitoring; Intellia's fix is patient selection — genotyping enrollees, taking it to the FDA, building a risk-mitigation/labeling strategy. Candid so-what for Aletira, with the same discipline held all week: what this is NOT — nex-Z is LNP-delivered in-vivo CRISPR (not AAV) and the driver is a host immune-genetic factor (a patient's HLA type), NOT the expression-selectivity axis SELEXON governs — so it is not "SELEXON would have prevented this," and pitching it that way hands a sharp investor a seam. What it IS: one more brick in the 2026 wall that systemic-genetic-medicine toxicity is driven by host immune/genetic factors and the bar for controlling it keeps ratcheting up (the same rising-safety-bar tailwind under the raise and the Aug 23 Ultragenyx decision). New, additive point: Intellia's answer to a safety problem is precision — restrict the therapy to the right patients by genotype — which rhymes with Aletira's instinct to restrict expression to the right cells; the field is visibly converging on precision/selection as the path to approvable safety, the exact climate a selectivity platform wants to raise into. Framing for Hemaka/Eddie ahead of Geoff Lynn's raise: not a validation claim, a climate read. Board (one line each, no re-coverage of yesterday's two dates): Aletira — Ultragenyx AAV decision Aug 23 is the live safety-bar read, sharpened by this Intellia disclosure; 1104health — HHS OIG anti-kickback safe-harbor comment deadline Aug 24 needs a document from Rose Wang's team now; GPR52 NewCo (Lieber + Third Rock) — Nxera's rival still unsold, reprieve holds; RNA translation-activator platform — CAMP4 first-in-human and fully funded, the differentiated wedge (mechanism/delivery/disease set) is this week's live task; aSKY — nothing new since Revolution Medicines' quarter, mutant-RAS bar keeps rising; gut-restricted GCPII/IBD — TL1A crowding, differentiate on the enteric-pain axis. Source: BioSpace, August 7, 2026 (on Intellia's August 6 Q2 update). Link: https://www.biospace.com/drug-development/intellia-finds-genetic-suspect-for-liver-safety-signals-with-attr-gene-therapy

Portfolio Watch — a quiet-tape week-ahead: two hard August dates (the Ultragenyx AAV decision, the OIG comment deadline) that need work from us now

Sat, 08 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 8, 2026 — a thirteenth straight quiet direct-catalyst day (no readout, financing, deal, or filing landed on a Blackbird company this week). The only genuinely in-window items were already aired and are not re-covered: the two more pediatric gene-editing deaths in China (yesterday's lead), Revolution Medicines' Q2 blowout (Wednesday's lead), and the AstraZeneca–Bristol Myers merger rumor (knocked down mid-week, story settled). Rather than pad, a week-ahead on the two hard calendar dates that need action now. Lead (lands on the flagship Aletira): the FDA decision due August 23 on Ultragenyx's adeno-associated-virus gene therapy DTX401 for glycogen storage disease type Ia (PDUFA date August 23, 2026) — reframed, because this will be the first FDA verdict on a systemic AAV gene therapy to land after the safety picture darkened (Sarepta Elevidys deaths, then this week's two China gene-editing deaths, both immune reactions driven by capsid dose and biodistribution). So August 23 is a live test of how far the agency's risk tolerance has moved: a clean approval says the bar is efficacy-plus-manageable-safety; an approval bolted to a tight label/boxed warning/heavy post-marketing commitment — or a delay/CRL — says the precision-and-safety bar has ratcheted up, which is the tailwind under Aletira's SELEXON story. Standing caveat kept: SELEXON governs expression selectivity (which cell type reads out the transgene), not capsid dose or biodistribution — the axis all those deaths turned on — so it's thesis validation and a fundraising talking point, not a claim it would have prevented them. Action for Hemaka/Eddie: read the August 23 outcome as a timing input for sequencing Geoff Lynn's raise (a visibly tightening safety bar is when a precision-selectivity story sells best). Second date, August 24 (needs a filing from us): the HHS Office of Inspector General's request for information on the anti-kickback statute and beneficiary-inducement rules in the clinical-trial context closes to comment. Its headline is remuneration to trial participants, but it reopens the whole anti-kickback safe-harbor framework for trial-related payments — the framework that governs 1104health's revenue mechanism (pharma-funded subsidies compensating community oncologists for previously unreimbursed trial-coordination work). It is the open door to put the community-oncology coordination model in front of regulators while they are actively rethinking those safe harbors; the comment Rose Wang's team files should make the case for a clear line between fair-market-value reimbursement for real coordination labor and anything resembling an inducement to enroll. Sixteen days out with a document to draft = the work starts this weekend. Standing board (no re-coverage): GPR52 NewCo (Lieber + Third Rock) — Nxera's rival GPR52 agonist still unsold, reprieve holds; RNA translation-activator platform — CAMP4 first-in-human and fully funded, so Avi and the platform leads defining the differentiated wedge (mechanism, delivery, disease set) is this week's live task; aSKY — nothing new since Wednesday's Revolution Medicines quarter, the mutant-RAS bar keeps rising; gut-restricted GCPII/IBD — TL1A crowding, differentiate on the enteric-pain axis. Sources: Ultragenyx (GlobeNewswire, Feb 23, 2026, PDUFA Aug 23); HHS OIG RFI (Federal Register, June 24, 2026, comments due Aug 24). Link: https://www.globenewswire.com/news-release/2026/02/23/3242633/20739/en/Ultragenyx-Announces-U-S-FDA-Acceptance-and-Priority-Review-of-the-Biologics-License-Application-BLA-for-DTX401-AAV-Gene-Therapy-for-Glycogen-Storage-Disease-Type-Ia-GSDIa.html

Portfolio Watch — the gene-therapy safety reckoning broadens (two more pediatric gene-editing deaths in China); a candid read on where it does and doesn’t help Aletira

Fri, 07 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 7, 2026 — a twelfth straight quiet direct-catalyst day on Blackbird companies (no readout, financing, deal, or filing landed on a program this week); the two threads that moved earlier are settled and not re-covered (CAMP4’s ~$50M second close is done; the AstraZeneca–Bristol Myers merger rumor was knocked down Wednesday with no follow-on). Lead (in-window, lands on the flagship Aletira as a landscape read, not a catalyst): the gene-therapy safety picture that has validated the precision thesis all year darkened this week as two more pediatric gene-editing deaths in China surfaced. HuidaGene Therapeutics (Shanghai) disclosed August 5 that a boy in its HG302 CRISPR Duchenne trial died last August of acute respiratory distress during a severe immune reaction after a high dose (three other dosed boys reportedly stable); separately, a 6-year-old girl died in March 2025, seven days after an investigator-led team infused trillions of viral particles (dual AAV9) into her spinal fluid to base-edit a brain gene — cause of death a severe immune reaction with thrombotic microangiopathy, judged “definitely related” by the hospital ethics board, and kept out of a subsequent Nature paper (now facing retraction calls). Candid so-what for Aletira: both deaths are immune-mediated and driven by capsid dose and biodistribution — how much virus is injected and where it goes — which is exactly the axis SELEXON does NOT govern (SELEXON controls expression selectivity, which cell type reads out the transgene, not capsid dose/biodistribution), so do not pitch these deaths as direct validation; a sharp investor will take that seam apart. What is legitimately ours: expression selectivity as part of the precision/therapeutic-index toolkit the field now demands (avoiding wrong-tissue readout that feeds immunogenicity; potentially reaching effect at lower vector dose). Second, genuinely new angle: the girl’s case centers outrage on gene therapy in a non-fatal condition — and Aletira’s lead indication, hereditary hearing loss, is also non-fatal, so the risk-benefit/safety bar for non-lethal pediatric gene therapy is rising. Action for Hemaka/Eddie ahead of Geoff Lynn’s raise: sharpen the pitch around the axis SELEXON actually controls plus the dose/therapeutic-index story, and pressure-test the hearing-loss indication strategy against a higher non-fatal safety bar. Related underwriting datapoint: Aurora Therapeutics (Jennifer Doudna co-founded personalized gene-editing startup scaling Baby-KJ-style treatments) killed its lead program and cut staff August 4, ~7 months after launch, citing competition, IP hurdles, and manufacturing — a reminder that clearing a heroic one-patient bar is not a scalable, commercializable business (failure modes increasingly economic/CMC, not just biological). Standing board (no re-coverage): 1104health — HHS OIG anti-kickback safe-harbor RFI on paying trial participants closes August 24 (get a comment/position in for Rose Wang’s team); gene-therapy FDA calendar — Ultragenyx’s rare-metabolic-disease decision due August 23 is a live read on AAV safety posture; GPR52 NewCo (Lieber + Third Rock) — Nxera’s rival GPR52 agonist still unsold, reprieve holds; gut-restricted GCPII/IBD — TL1A crowding, differentiate on the enteric-pain axis. Source: STAT News, August 5, 2026. Link: https://www.statnews.com/2026/08/05/china-clinical-trial-fatality-safety-questions-investigator-led-studies/

Portfolio Watch — RevMed's blowout Q2 raises the bar for aSKY; the AZ–Bristol Myers merger rumor is denied, in our favor

Thu, 06 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 6, 2026 — the direct-catalyst tape stays quiet on Blackbird companies, but two tracked threads now have real content. Lead (in-window, the key aSKY competitive read): Revolution Medicines reported Q2 after the close August 5, and it is a strong quarter for the mutant-RAS reference-setter. On daraxonrasib (multi-selective RAS(ON) inhibitor), the FDA accepted the pancreatic-cancer NDA into the Commissioner's National Priority Voucher pilot and the EMA began a phased review; the May expanded-access program has, within about three weeks, reached physicians representing more than 2,000 pancreatic-cancer patients across nearly every state, and the company declared U.S. commercial launch readiness — no longer a clinical story, a launch. Beyond pancreatic: FDA breakthrough-therapy designation for daraxonrasib in previously-treated RAS-mutant (non-G12C) lung cancer, pivotal lung enrollment completing this year with a 2027 readout, and first-line lung combination data of 85% confirmed response for elironrasib (G12C) and 82% for zoldonrasib (G12D) with clean liver safety — driving two new first-line Phase 3s. With $3.9 billion in cash, RevMed is fully funded to march the entire RAS(ON) franchise across every mutation and line into the market at once. So-what for aSKY (Blackbird's UMB KRAS-directed antibody, mechanism undisclosed, AVS Bio partnership): this is not a competitor's bad day, it is the reference-setter's exceptional quarter — every quarter the small-molecule lane de-risks and commercializes across G12C/G12D/G12V, the bar aSKY's eventual mechanism disclosure must clear rises and the cost of staying undisclosed rises with it. A low-80s first-line lung response rate is now the ambient reference a differentiated modality gets measured against. Action for Eddie/aSKY: read the transcript as competitive intelligence, and press on when and how aSKY articulates what an antibody does that a small molecule cannot — "the way at RAS the small molecules can't reach," not "another way at RAS." Second thread (resolves in our favor): the AstraZeneca–Bristol Myers merger rumor that anchored Monday's and Tuesday's briefs has effectively been knocked down — a senior source told Reuters August 5 there are no discussions and "there never was a deal to be done"; AstraZeneca shares recovered their ~9% slide. For our read this is a positive: with Bristol Myers staying independent, it keeps Cobenfy and stays on the board as an independent neuro-psychiatric buyer for the Lieber + Third Rock GPR52 NewCo — the thinner-exit-optionality worry from Tuesday lifts, and the near-term oncology-consolidation overhang on aSKY's partnering conversation eases. Caveat: a source-level denial isn't a press release and the strategic logic of an oncology mega-combination is now in the open, so don't file it away permanently. Standing board (no re-coverage): CAMP4's ~$50M second close was yesterday's lead — the clock it started is running, so Avi/platform leads' differentiation wedge (translation-step mechanism, delivery beyond intrathecal, disease set where CAMP4 is absent) stays this week's assignment; 1104health — the HHS Inspector General's anti-kickback safe-harbor RFI on paying trial participants closes August 24, worth a comment or position paper for Rose Wang's team; Aletira — no fresh news, gene-therapy FDA calendar (Ultragenyx GSD type 1-A, Aug 23) keeps precision/SELEXON topical (caveat: SELEXON governs expression selectivity, not capsid biodistribution); gut-restricted GCPII/IBD — TL1A crowding (tulisokibart, duvakitug), differentiate on the enteric-pain axis. Source: Revolution Medicines Q2 2026 press release, August 5, 2026. Link: https://www.biospace.com/press-releases/revolution-medicines-reports-second-quarter-2026-financial-results-and-update-on-corporate-progress

Portfolio Watch — CAMP4's ~$50M translation-activator tranche closes; RevMed reports tonight; quiet-catalyst day ten

Wed, 05 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 5, 2026 — a tenth straight quiet direct-catalyst day (no readout, financing, deal, or filing landed squarely on a Blackbird company), but the week's tracked financing finally has its paperwork. Lead (in-window, confirmed): CAMP4 Therapeutics closed the second tranche of its private placement on August 4 — approximately $50.1 million gross (about 10.8 million common shares at $1.53, roughly $22 million more in pre-funded warrants at essentially the same price, plus a small insider slug), the pre-committed September-2025 syndicate (Coastlands lead, Janus Henderson, Vivo, 5AM, Balyasny, Adage, Trails Edge, CURE SYNGAP1; Leerink lead placement agent), mechanically triggered by the July 27 Australian clearance of CMP-002 into a Phase 1/2 in SYNGAP1-related disorder. Candid read: this is a pre-committed insider milestone tranche at a sub-$2 price, not a hot new round at a step-up, so temper the validation — but it takes money off the constraints list for the named independent competitor in the RNA translation-activator (turn-the-protein-up) lane our platform lives in, now first-in-human and fully funded into the clinic. So-what: the differentiation wedge Avi and the platform leads own — translation-level mechanism vs CAMP4's transcriptional regRNA-ASO; delivery beyond intrathecal into liver/muscle/periphery; disease set where a translation-step edge is real and CAMP4 is absent — is now on a live clock, write it down this week. Near-term calendar: Revolution Medicines reports Q2 after the close today, the single most decision-relevant listen for aSKY's mutant-RAS competitive map (daraxonrasib pancreatic review timeline under the National Priority Voucher pilot, lung-cancer pace/combinations, first-line ambitions) — a calendar event, not news, read as competitive intelligence. AstraZeneca–Bristol Myers merger rumor: no change — no confirmation or denial since the Aug 2–3 Financial Times report and market reaction, no fresh Aug 4–5 reporting; yesterday's three-thread read (thinner exit optionality for the GPR52 NewCo if Bristol Myers/Cobenfy folds into oncology-heavy AstraZeneca; the combined oncology giant as aSKY's biggest potential buyer/competitor; the macro concentration of the acquirer set) stands — watch, don't chase. Standing board: GPR52 neuropsychiatric NewCo (Lieber + Third Rock) — Nxera's rival GPR52 agonist still unsold, reprieve holds; 1104health — concrete forward marker: the HHS Inspector General's RFI on anti-kickback safe harbors for clinical-trial participant remuneration has an August 24 comment deadline, the exact policy surface that governs paying physicians for trial coordination, worth a comment or position paper; Aletira — no fresh news, gene-therapy FDA calendar (Ultragenyx GSD type 1-A, Aug 23) keeps precision/SELEXON topical (caveat: SELEXON governs expression selectivity, not capsid biodistribution); gut-restricted GCPII/IBD — TL1A crowding (tulisokibart, duvakitug), differentiate on the enteric-pain axis. Source: CAMP4 Therapeutics press release, August 4, 2026. Link: https://www.biospace.com/press-releases/camp4-therapeutics-announces-second-closing-of-100-million-private-placement

Portfolio Watch — the AstraZeneca–Bristol Myers merger rumor lands on our schizophrenia buyer map; quiet-catalyst day nine

Tue, 04 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 4, 2026 — a ninth straight quiet direct-catalyst day (no readout, financing, deal, or filing landed on a Blackbird asset), but with one genuinely fresh in-window development worth leading on as a landscape signal, not a catalyst. Lead: the Financial Times reported (Aug 2) that AstraZeneca is in early-stage talks to merge with Bristol Myers Squibb — a roughly $400 billion combination, among the largest in industry history — talks the companies have held on and off for months, may not proceed, and that drew immediate antitrust flags over overlapping oncology (AstraZeneca oncology alone was ~$25 billion in 2025). So-what for Blackbird: (1) Sharpest thread — the Lieber Institute + Third Rock GPR52 schizophrenia NewCo. Bristol Myers owns Cobenfy, the muscarinic (non-dopaminergic) antipsychotic and the marquee competitive reference for a GPR52 asset, and is one of the short-list strategic buyers we track for this space; folding it into an oncology/rare-disease-heavy AstraZeneca with no psychiatry franchise makes the combined entity’s CNS priorities uncertain and shrinks the pool of independent neuro-psych buyers — thinner exit optionality long term, though a distracted integrating acquirer may also extend the reprieve on Nxera’s still-unsold GPR52 rival. (2) aSKY — a combined AstraZeneca-Bristol Myers is a $25B+ oncology giant, the single largest potential partner, acquirer, and competitor in aSKY’s KRAS solid-tumor space; megamergers freeze BD for a year-plus and shed deprioritized pipeline (near-term buyer-pool concentration, longer-term a sourcing surface for a studio). (3) Macro — consolidation at the top concentrates the acquirer set every Blackbird spinout exits into, making each exit more binary, while also loosening assets and talent. Candid caveat: this is a reported early-stage talk, not a signed deal — watch it, don’t chase it. Standing board: CAMP4 second private placement was expected to close on/about Aug 3, no confirming SEC filing yet as of this morning (watch item, no re-litigation) — the fully-funded RNA translation-activator pace-setter, Avi’s wedge work is this week; Revolution Medicines Q2 call Wed Aug 5 after close is the near-term aSKY mutant-RAS competitive input, now doubly worth watching against the oncology-consolidation backdrop (calendar event, not news); Aletira — no fresh news, precision/SELEXON thesis topical, Ultragenyx GSD-Ia decision Aug 23 (caveat: selectivity, not biodistribution); 1104health — no news, Operation TrialBlazer payment-reform/OIG compliance watch stands; gut-restricted GCPII/IBD — TL1A crowding, differentiate on the enteric-pain axis. Source: CNBC/Financial Times, August 2, 2026. Link: https://www.cnbc.com/2026/08/02/astrazeneca-and-bristol-myers-squibb-mull-400-billion-deal-report-.html

Portfolio Watch — CAMP4’s ~$50M translation-activator financing closes today; the concrete wedge, and quiet day eight

Mon, 03 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 3, 2026 — the follow-through on Sunday’s live thread, and an eighth straight quiet direct-catalyst day (no readout, financing, deal, or filing landed on a Blackbird asset). Lead (in-window, lands today): CAMP4 Therapeutics’ second private placement — up to $50.1 million gross for roughly 32.7 million shares and pre-funded warrants — is expected to close on or about today, subject to customary conditions, triggered by the July 27 clearance of CMP-002 into a first-in-human SYNGAP1 trial in Australia. Leerink is lead placement agent (Piper Sandler, Cantor, Wedbush co-agents); committed investors include Coastlands, Janus Henderson, Vivo, 5AM, Balyasny, Adage, Trails Edge, and CURE SYNGAP1. Read: CAMP4 is the named independent competitor in the RNA translation-activator (“turn-the-protein-up”) lane our platform lives in, and it is now fully funded and in the clinic — money off its list of constraints. Not a threat to a specific Blackbird asset (no competing SYNGAP1 program) but a pace-setter, so the differentiation task for Avi and the platform leads is now this week’s work, along three concrete axes: mechanism (translation-level control vs CAMP4’s transcriptional regRNA-ASO — different chemistry and IP, with a dose-control/reversibility edge on some targets), delivery (peripheral, liver, and muscle white space vs CAMP4’s intrathecal CNS reach), and disease set (plant a flag where turning protein up at the translation step has a biological edge and CAMP4 is absent, not on SYNGAP1 or the marquee names they’ve claimed). Near-term calendar: Revolution Medicines’ second-quarter call on Wednesday, August 5, a mutant-RAS competitive-intelligence input for aSKY’s undisclosed KRAS-directed antibody frame — not fresh news. Standing board unchanged: Neuropsychiatric NewCo (Lieber + Third Rock) — Nxera’s GPR52 rival still unsold, reprieve holds; Aletira — no fresh news, precision/cell-selectivity thesis stays topical against the Sarepta Elevidys safety backdrop (caveat: SELEXON governs expression selectivity, not capsid biodistribution); 1104health — keep the Operation TrialBlazer payment-reform/safe-harbor story airtight; gut-restricted GCPII/IBD — TL1A crowding means differentiate on the enteric-pain axis. Source: CAMP4 Therapeutics Form 8-K (SEC), filed July 27, 2026. Link: https://www.sec.gov/Archives/edgar/data/0001736730/000162828026049661/camp-20260727.htm

Sourcing Radar — U-M-B’s Rassool lab: a STING agonist plus decitabine turns p53-mutant AML into an immunotherapy target

Sun, 02 Aug 2026 12:00:00 +0000

The Blackbird Brief Sourcing Radar for August 2, 2026. This week’s lead is a fresh, full-text-readable bioRxiv preprint from Feyruz Rassool’s group at the University of Maryland School of Medicine / Greenebaum Comprehensive Cancer Center (posted 2026-07-28), with Johns Hopkins epigenetics co-authors Stephen Baylin and Michael Topper. TP53-mutated acute myeloid leukemia — ~10% of de novo and 25% of therapy-related AML, with no established immunotherapy and poor outcomes on chemo, transplant, and even venetoclax+decitabine — carries a hidden vulnerability: losing p53 makes leukemia cells MORE sensitive to a STING agonist. Using Curadev Pharma’s next-generation, human-specific, IV-dosable STING agonist C92, the team shows potentiated STING activation and cytokine release in p53-mutant vs wild-type cells; adding the generic DNA-methyltransferase inhibitor decitabine synergizes (unsilenced repetitive elements drive a viral-mimicry interferon response), and killing proceeds by inflammatory necroptosis through a ZNFX1/ZBP1–RIPK3–MLKL axis rather than apoptosis. In humanized mice the combination cut leukemia burden and pulled cytotoxic T cells into the tumor. So-what for Blackbird: the biology is strong and the p53-mutant biomarker is exactly the patient-selection story the STING-agonist class has lacked — but the IP is a use/biomarker/combination play, not a composition-of-matter spin-out: decitabine is generic, C92 belongs to Curadev, and the authors disclose no filed patent. Watch-and-partner, not found-and-lead: Eddie to probe whether UM Ventures will file on the method-and-biomarker and whether the home is a Curadev co-development deal; Avi/Yixuan to test what is ownable and defensible when you don’t control the molecule. The rest of the Hopkins/U-M-B/Lieber week was thin — Kwon-lab SF3B2 cardiomyocyte splicing biology (no asset), Hopkins tool/methods preprints (Neuropixels probe, pangenome patching, real-time MRI AI), and a Krantz metabolic-gating perspective. Paper link: https://www.biorxiv.org/content/10.64898/2026.07.27.741029v1

Portfolio Watch — CAMP4’s ~$50M close lands on the RNA translation-activator lane; quiet-tape day seven

Sun, 02 Aug 2026 13:00:00 +0000

The Blackbird Brief Portfolio Watch for August 2, 2026 — a quiet-tape edition: for the seventh straight day no direct catalyst (readout, financing, deal, filing) landed on a Blackbird company. The one genuinely in-window event bears on the RNA translation-activator platform: CAMP4 Therapeutics’ second private-placement tranche of up to $50 million (Janus Henderson, Vivo, 5AM, Balyasny, Adage, Coastlands, Trails Edge, CURE SYNGAP1) is expected to close within five business days of its July 27 SYNGAP1 FIH-clearance announcement — i.e., any day now. CAMP4 is the named independent competitor in the “turn-the-protein-up” modality (RNA to raise gene expression for haploinsufficiency/LOF disease); over two weeks it has cleared a regulator into humans, framed >1,200 addressable disorders, and now funded the program deep into the clinic. Not a threat to a specific Blackbird asset (no competing SYNGAP1 program) but a pace-setter starting a live clock — Avi and the platform leads’ task to define the differentiated wedge (targeting mechanism, delivery beyond intrathecal, disease set) goes from important to urgent. Standing board (no change): aSKY — Revolution Medicines’ Q2 call Wed Aug 5 (after close) is a competitive-intel input on the mutant-RAS class, not fresh news; Neuropsychiatric NewCo (Lieber + Third Rock) — Nxera’s GPR52 rival still unsold/unpartnered, modest reprieve holds; Aletira — no fresh news, August gene-therapy decision calendar keeps precision/SELEXON topical post-Elevidys (caveat: selectivity, not biodistribution); 1104health — late-July FDA eligibility guidances remain a demand tailwind, keep the Operation TrialBlazer payment-reform/safe-harbor story airtight; gut-restricted GCPII/IBD — TL1A crowding (tulisokibart Ph3 UC) means differentiate on the enteric-pain axis. Source: CAMP4 GlobeNewswire, 2026-07-27. Link: https://www.globenewswire.com/news-release/2026/07/27/3333378/0/en/CAMP4-Therapeutics-Secures-Australian-Regulatory-Clearance-to-Initiate-First-in-Human-Clinical-Trial-of-CMP-002-in-Patients-With-SYNGAP1-Related-Disorder.html

Portfolio Watch — a sixth straight quiet direct-catalyst day (no Blackbird program had a readout, financing, deal, or filing land), so a week-ahead brief rather than a manufactured story. The one must-watch is Wednesday: Revolution Medicines' second-quarter call on August 5, a competitive-intelligence input for aSKY's RAS frame — Revolution Medicines is de-risking mutant-RAS oncology fast (daraxonrasib NDA accepted in metastatic pancreatic cancer into the FDA National Priority Voucher pilot on a roughly doubled survival readout; G12D-selective zoldonrasib above 50% response in second-line KRAS G12D non-small-cell lung cancer), so listen for the pancreatic review timeline, the lung-cancer pace and combinations, and first-line ambitions. Aletira gets a landscape read, not a catalyst: the August gene-therapy decision calendar (Ultragenyx's glycogen storage disease type Ia decision due August 23, Sanfilippo behind it) keeps precision and cell-type control topical against a field still absorbing the Sarepta Elevidys AAV safety collapse — the tailwind under SELEXON's Series-A story (with the standing caveat that SELEXON governs expression selectivity, not capsid biodistribution). Plus four unchanged-but-live watch items: the Lieber/Third Rock GPR52 NewCo (Nxera's phase-2-ready rival still unsold; a Q3 buyer resets the clock), the RNA translation-activator platform (CAMP4 first-in-human clock running), 1104health (Operation TrialBlazer payment-track/anti-kickback compliance watch), and the gut-restricted GCPII program for IBD (TL1A crowding).

Sat, 01 Aug 2026 13:00:00 +0000

Portfolio Watch for Saturday, August 1, 2026. A genuinely quiet-tape, week-ahead edition. For the sixth trading day running, no Blackbird company or program had a readout, financing, deal, or filing land in the window; a full Stream-A sweep (industry press plus every program's target/modality/company queries, plus a general biotech scan) confirmed the fresh items were all off-thesis or already covered. So instead of dressing up a generic headline, a map of what actually matters for the portfolio in the week ahead, program by program, plus the standing watch items. **1. THE ONE MUST-WATCH (forward marker) — Revolution Medicines' Q2 call, Wednesday, August 5, after the close, is the most decision-relevant near-term event for aSKY.** Revolution Medicines runs the leading small-molecule RAS(ON) franchise and is setting the bar in the mutant-RAS oncology space aSKY has to define itself against: lead drug daraxonrasib (multi-selective RAS(ON) inhibitor) had its NDA accepted July 22 in previously-treated metastatic pancreatic cancer — into the FDA Commissioner's National Priority Voucher pilot — on a phase 3 readout that roughly doubled median overall survival vs chemotherapy; the G12D-selective drug zoldonrasib shows greater-than-50% response in second-line KRAS G12D non-small-cell lung cancer with breakthrough designation; the broader phase 3 program reaches into lung cancer. **So-what:** aSKY's disclosed frame is a KRAS-directed antibody program (via the AVS Bio engineering partnership) — a differentiated modality against a field of small molecules, but the mechanism is still undisclosed, and every de-risking datapoint out of Revolution Medicines raises the bar aSKY's eventual disclosure must clear. **Action:** Eddie + aSKY team treat Wednesday's call as a competitive-map input — pancreatic review timeline under the priority voucher, lung-cancer pace/combinations, first-line ambitions. https://www.globenewswire.com/news-release/2026/07/29/3335558/0/en/revolution-medicines-to-report-financial-results-for-second-quarter-2026-after-market-close-on-august-5-2026.html ; https://www.globenewswire.com/news-release/2026/07/22/3331714/0/en/Revolution-Medicines-New-Drug-Application-for-Daraxonrasib-Accepted-for-Review-by-U-S-FDA-for-Previously-Treated-Metastatic-Pancreatic-Cancer.html **2. Aletira — landscape read, not a catalyst (no fresh company news).** The August gene-therapy decision calendar keeps the precision thesis topical: Ultragenyx has an FDA decision due August 23 on its AAV gene therapy for glycogen storage disease type Ia, and another AAV decision in September for Sanfilippo. Zoom out and the 2026 story is that the A-A-V hurdle has moved from efficacy to safety and manufacturing consistency after the Sarepta Elevidys collapse (liver-failure deaths, boxed warning, revoked platform designation) — which is precisely the tailwind under Aletira's Series-A story, since SELEXON (alternative-splicing selectivity licensed from Johns Hopkins) sells restriction of *where* a transgene turns on into a market that just learned why that matters. **Standing caveat (do not over-claim):** SELEXON governs expression selectivity, not capsid biodistribution, and much of Elevidys tox was dose/biodistribution-driven — thesis validation and a fundraising talking point, not a claim SELEXON would have prevented Elevidys. Watch item; Hemaka's pre-raise touchpoint with Geoff Lynn stands. https://www.globenewswire.com/news-release/2026/02/23/3242633/20739/en/Ultragenyx-Announces-U-S-FDA-Acceptance-and-Priority-Review-of-the-Biologics-License-Application-BLA-for-DTX401-AAV-Gene-Therapy-for-Glycogen-Storage-Disease-Type-Ia-GSDIa.html ; https://biobuzz.io/news/aletira-therapeutics-becomes-blackbird-labs-first-fully-incubated-spinout-and-proof-point-for-baltimores-biotech-thesis/ **3. GPR52 / schizophrenia NewCo (Lieber Institute + Third Rock) — no change.** Nxera's phase-2-ready GPR52 agonist (the only other one in industry) remains unsold this window, so the modest timing reprieve holds ahead of development-candidate declaration later this year; a Q3 pickup by Otsuka / Bristol Myers (via Karuna) / Neurocrine / AbbVie (via Cerevel) resets the differentiation clock. Sharpen the best-in-class chemistry story now. https://www.globenewswire.com/news-release/2025/12/18/3208254/0/en/Nxera-Pharma-to-Regain-Full-Rights-to-GPR52-Agonist-Program-for-Schizophrenia.html **4. RNA translation-activator platform — no change since 2026-07-28.** CAMP4's CMP-002 first-in-human clock is running (validation of the turn-the-protein-up modality, and a pace-setter). Action stands: Avi + platform leads define the defensible wedge (targeting mechanism / delivery / disease set). **5. 1104health — watch-item stands (from 2026-07-29).** No new FDA guidance this week; Rose Wang's team should keep reading the Operation TrialBlazer payment-reform track (HHS Office of Inspector General at the table) as closely as the eligibility guidances — that workstream governs whether paying physicians for trial-coordination work stays clean (anti-kickback / safe-harbor) as they scale. **6. Gut-restricted GCPII program for IBD — no change.** The TL1A wave keeps crowding the inflammatory-bowel-disease space (tulisokibart through phase 3 in ulcerative colitis; the Sanofi-Teva antibody positive mid-stage), so differentiation must keep leading on the enteric-pain axis. https://www.fiercebiotech.com/biotech/mercks-anti-tl1a-antibody-prometheus-buyout-passes-ph-3-test-ulcerative-colitis

Portfolio Watch — a genuinely quiet direct-catalyst tape (no Blackbird program had a readout, financing, deal, or filing land this week), so a short, honest brief anchored on the one near-term event that actually reprices a portfolio thesis: Revolution Medicines' second-quarter call next Tuesday, August 5, and what it means for aSKY. Revolution Medicines is setting the bar in mutant-RAS oncology — daraxonrasib's NDA was accepted July 22 in previously-treated metastatic pancreatic cancer (into the FDA National Priority Voucher pilot) on an unprecedented phase 3 survival readout, with G12D-selective zoldonrasib showing >50% response in second-line KRAS G12D non-small-cell lung cancer — so the field aSKY's undisclosed KRAS-frame antibody program must differentiate against keeps de-risking, and the bar keeps rising. Listen Tuesday for the pancreatic review timeline, the non-small-cell lung cancer pace/combination plans, and first-line ambitions. Plus tight no-change watch-items: the Lieber/Third Rock GPR52 NewCo (Nxera's phase-2-ready rival still unsold; a Q3 buyer resets the clock), the RNA translation-activator platform (CAMP4 first-in-human clock running), Aletira (no fresh catalyst; AAV-safety field backdrop stays live), and 1104health (Operation TrialBlazer payment-track/anti-kickback compliance watch stands).

Fri, 31 Jul 2026 13:00:00 +0000

Portfolio Watch for Friday, July 31, 2026. A genuinely quiet direct-catalyst tape: across the past few days, no Blackbird company had a readout, financing, deal, or filing land in the window, and a full Stream-A sweep (industry press plus every program's target/modality/company queries) confirmed the fresh items were all out of window or already covered. So a short, honest brief anchored on the one near-term event that actually moves a portfolio thesis, plus a tight status pass. **1. LEAD (forward marker, not a landed catalyst) — Revolution Medicines' Q2 call next Tuesday, August 5, is the most decision-relevant near-term event for aSKY.** Revolution Medicines announced (July 29) it will report second-quarter results after market close on August 5. It runs the leading small-molecule RAS(ON) franchise and is setting the bar in the mutant-RAS oncology space aSKY has to define itself against: lead drug daraxonrasib (multi-selective RAS(ON) inhibitor) had its NDA accepted July 22 in previously-treated metastatic pancreatic cancer — into the FDA Commissioner's National Priority Voucher pilot (can pull review forward) — on a phase 3 readout that roughly doubled median overall survival vs chemotherapy (13.2 vs 6.7 months, HR 0.40); the G12D-selective drug zoldonrasib shows >50% response in second-line KRAS G12D non-small-cell lung cancer with breakthrough designation; and the broader phase 3 program reaches into lung cancer. **So-what for Blackbird:** aSKY's disclosed frame is a KRAS-directed program built on antibody engineering (via the AVS Bio partnership) — a different modality from Revolution Medicines' small molecules, with the specific mechanism still undisclosed. Different modality is how you differentiate in a crowded target space, but only if the story is sharp — and every de-risking datapoint out of Revolution Medicines raises the bar aSKY's eventual disclosure must clear and makes the undisclosed-mechanism gap more expensive each quarter. **Action:** Eddie + aSKY team treat Tuesday's call as a competitive-map input — listen for the pancreatic review timeline under the priority voucher, the non-small-cell lung cancer pace/combination plans, and first-line ambitions. https://www.globenewswire.com/news-release/2026/07/29/3335558/0/en/revolution-medicines-to-report-financial-results-for-second-quarter-2026-after-market-close-on-august-5-2026.html ; https://www.globenewswire.com/news-release/2026/07/22/3331714/0/en/Revolution-Medicines-New-Drug-Application-for-Daraxonrasib-Accepted-for-Review-by-U-S-FDA-for-Previously-Treated-Metastatic-Pancreatic-Cancer.html ; https://www.umbiopark.com/biopark-companies/asky-therapeutics **2. GPR52 / schizophrenia NewCo (Lieber Institute + Third Rock) — no change.** Nxera's phase-2-ready GPR52 agonist (the only other one in industry) remains unsold this window, so the modest timing reprieve holds ahead of development-candidate declaration later this year; a Q3 pickup by Otsuka / Bristol Myers (via Karuna) / Neurocrine / AbbVie (via Cerevel) would reset the differentiation clock (12–24 months behind a same-mechanism rival). The watch is a deal, and it hasn't happened. https://www.globenewswire.com/news-release/2025/12/18/3208254/0/en/Nxera-Pharma-to-Regain-Full-Rights-to-GPR52-Agonist-Program-for-Schizophrenia.html **3. RNA translation-activator platform — no change since 2026-07-28.** The first-in-human clock that started Monday with CAMP4's CMP-002 clearance is now running (validation of the turn-the-protein-up modality, and a pace-setter). Action stands: Avi + platform leads define the defensible wedge (targeting mechanism / delivery / disease set) before the lead widens. https://www.globenewswire.com/news-release/2026/07/27/3333378/0/en/CAMP4-Therapeutics-Secures-Australian-Regulatory-Clearance-to-Initiate-First-in-Human-Clinical-Trial-of-CMP-002-in-Patients-With-SYNGAP1-Related-Disorder.html **4. Aletira — no fresh catalyst since the 2026-07-30 positioning read (not re-run).** Backdrop only: the AAV gene-therapy safety story stays live, with a run of gene-therapy regulatory decisions over the next few weeks — each keeps precision / cell-type control in the conversation SELEXON wants to be having. **5. 1104health — watch-item stands (from 2026-07-29).** No new FDA guidance this week; Rose Wang's team should keep reading the Operation TrialBlazer payment-reform track (HHS Office of Inspector General at the table) as closely as the eligibility guidances — that workstream governs whether paying physicians for trial-coordination work stays clean (anti-kickback / safe-harbor) as they scale.

Portfolio Watch — a quiet direct-catalyst tape, so a positioning read on the flagship: where Aletira sits after 2026's gene-therapy safety reckoning. The Sarepta Elevidys AAV liver-tox saga (boxed warning, non-ambulatory eligibility pulled, platform-technology designation revoked) has made untargeted, off-target transgene expression the field's most public liability — validating the premise of Aletira's SELEXON platform (alternative RNA splicing to restrict where a transgene expresses; licensed from Johns Hopkins), even though SELEXON governs expression selectivity, not capsid biodistribution. Meanwhile the lead-indication lane, hereditary hearing loss, has consolidated: Regeneron's otoferlin gene therapy Otarmeni is FDA-approved (and free in the US), Eli Lilly/Akouos' AK-OTOF is advancing, and Sensorion discontinued SENS-501 — so the easy single-cell-type target is spoken for and Aletira's differentiation lives in the harder targets where cell-type control is decisive. Plus forward markers: aSKY (Revolution Medicines Q2 call Aug 5), GPR52 NewCo (Nxera rival still unsold), RNA translation-activator platform (CAMP4 FIH clock running), GCPII/IBD, and the 1104health TrialBlazer payment-track watch.

Thu, 30 Jul 2026 13:00:00 +0000

Portfolio Watch for Thursday, July 30, 2026. A quiet day for direct portfolio-company catalysts — no readout, financing, deal, or filing landing on a Blackbird program in the window. Rather than manufacture a catalyst, today is a candid positioning read on the flagship, Aletira Therapeutics, which the recent daily run has talked around while leading on other programs. **1. POSITIONING READ (not a catalyst) — Aletira and the 2026 gene-therapy safety reckoning.** Aletira is Blackbird's proof point: the first fully incubated spinout (nonprofit grant to seed to a Blackbird BioHub bench). Its SELEXON platform, licensed from Johns Hopkins, uses alternative RNA splicing to restrict where a transgene is expressed — a bolt-on to a vector like an AAV that shuts off expression in unwanted cell types. Founding CEO Geoffrey M. Lynn (ex-Avidea, Vaccitech/Barinthus); lead indication signal is hereditary hearing loss. **Modality backdrop:** the defining gene-therapy story of 2026 has been the Sarepta Elevidys (AAV micro-dystrophin, Duchenne) safety reckoning — patient deaths from acute liver failure, an FDA boxed warning for acute liver injury, non-ambulatory eligibility pulled, platform-technology designation revoked, and a required ~200-patient postmarketing safety study. The through-line is systemic, off-target AAV exposure and hepatotoxicity. **So-what for Blackbird:** candidly, this is NOT "SELEXON would have prevented Elevidys" — SELEXON governs transgene expression selectivity, not where the capsid physically distributes, and much of the Elevidys liver signal is capsid-dose/biodistribution-driven (a separate axis). What it IS: validation of Aletira's core premise that untargeted expression is a real, now FDA-underlined liability of the modality — a tailwind for the thesis and a Series-A talking point as the field pivots toward precision. (Aletira's own framing, per BioBuzz: "off-target expression in the wrong tissues is one of the primary sources of toxicity, immunogenicity, and dose-limiting side effects in the field today.") **Lead-indication landscape (more double-edged):** hereditary hearing loss has gone from open frontier to a lane with a validated first target and real competition — Regeneron's otoferlin gene therapy Otarmeni (lunsotogene parvec) is FDA-approved as the first-and-only gene therapy for genetic hearing loss and is being provided free in the US (resetting the commercial reference point); Eli Lilly/Akouos' dual-vector AK-OTOF is advancing with striking early hearing restoration; and Sensorion discontinued SENS-501 earlier this summer. Implication: the easy small-gene, single-cell-type otoferlin target is spoken for; Aletira's real differentiation is in the harder monogenic hearing-loss forms where expressing in the wrong cochlear cell type is itself the problem. **Action:** Hemaka/Eddie touchpoint with Geoff Lynn's team before the next raise — where is SELEXON's expression control genuinely decisive vs. otoferlin-replacement, and does the seed-stage roadmap point at those differentiated targets rather than a me-too swing at the target Regeneron and Lilly already own. https://biobuzz.io/news/aletira-therapeutics-becomes-blackbird-labs-first-fully-incubated-spinout-and-proof-point-for-baltimores-biotech-thesis/ ; https://www.fda.gov/news-events/press-announcements/fda-approves-new-safety-warning-and-revised-indication-limits-use-elevidys-following-reports-fatal ; https://investor.regeneron.com/news-releases/news-release-details/otarmenitm-lunsotogene-parvec-cwha-approved-fda-first-and-only ; https://akouos.com/our-focus/ **2. aSKY Therapeutics — no change.** No new precision-NSCLC/RAS catalyst in the window; the field keeps hardening at the regulatory level. Next real read: Revolution Medicines' Q2 call on Aug 5 (non-small-cell combination plans, RASolve enrollment pace) — the differentiation bar aSKY's undisclosed KRAS-frame mechanism must clear keeps rising. Competitive-map refresh still on Eddie's list. **3. GPR52 / schizophrenia NewCo (Lieber Institute + Third Rock) — no change since 2026-07-27.** Nxera's Phase-2-ready GPR52 agonist remains unsold this window, so the modest timing reprieve holds; a Q3 pickup by Otsuka / BMS (via Karuna) / Neurocrine / AbbVie (via Cerevel) would reset the differentiation clock. **4. RNA translation-activator platform — no change since 2026-07-28.** The first-in-human clock started Monday with CAMP4's CMP-002 clearance is now running. Action stands: Avi + platform leads define the defensible wedge. **5. Gut-restricted GCPII program (JHU/Slusher) — quiet.** IBD competitive picture unchanged since the TL1A (tulisokibart) readouts; differentiation stays on the enteric-pain axis. **6. 1104health — watch-item stands (from 2026-07-29).** Read the Operation TrialBlazer payment/access-reform track (HHS OIG at the table) as closely as the eligibility guidances; airtight anti-kickback/safe-harbor compliance before national rollout.

Portfolio Watch — the FDA just widened the door into cancer trials, and it lands on eleven-oh-four health: the Oncology Center of Excellence finalized three eligibility guidances (performance status, laboratory values, washout periods/concomitant meds; posted July 27, Federal Register July 28) telling sponsors to drop scientifically unjustified exclusions — an execution step in HHS's Operation TrialBlazer. Fewer than 5% of cancer patients enroll in trials while >70% are willing; broadening eligibility expands the pool a community oncologist can match at the point of care = a demand-side tailwind for Blackbird BioVentures' clinician-embedded enrollment marketplace. Candid nuance: it doesn't fix the coordination-reimbursement gap (the core of the model), and the same initiative has the HHS Office of Inspector General at the table — the anti-kickback/safe-harbor watch on pharma-funded physician subsidies. Plus no-change notes on aSKY (RAS map), the GPR52 NewCo (Nxera rival still unsold), and the RNA translation-activator platform (CAMP4 FIH).

Wed, 29 Jul 2026 13:00:00 +0000

Portfolio Watch for Wednesday, July 29, 2026. A quiet day for direct portfolio-company catalysts; today's most important development is regulatory and lands on the portfolio's one non-therapeutic asset, eleven-oh-four health (1104health). **1. LEAD — FDA finalizes three cancer clinical-trial eligibility guidances (posted July 27, 2026; Federal Register July 28): a structural tailwind for 1104health's community-oncology enrollment marketplace.** The FDA Oncology Center of Excellence finalized three guidances instructing trial sponsors to stop excluding patients for reasons that aren't scientifically justified: (i) performance status — broaden beyond only the fittest patients on ECOG/Karnofsky scales (adults only); (ii) laboratory values — avoid blanket blood-count/lab cutoffs that screen out patients whose values reflect disease, age, or comorbidity rather than a real safety risk; (iii) washout periods and concomitant medications — justify inter-therapy waiting periods scientifically rather than via blanket exclusion. The FDA's framing statistic: fewer than 5% of cancer patients in treatment enroll in a trial while more than 70% say they'd be willing — and unnecessarily restrictive eligibility criteria are a named cause. These guidances are an execution step inside Operation TrialBlazer, the department-wide HHS initiative announced June 22, 2026 to rebuild U.S. clinical-research infrastructure (spanning FDA, NIH/NCI, ARPA-H, ONC, and — notably — the HHS Office of Inspector General). **So-what for Blackbird:** clean read is a demand-side tailwind. 1104health (Blackbird BioVentures portfolio; founder Rose Wang; JHTV-licensed know-how) pivoted to a clinician-embedded marketplace that surfaces trial options inside the community oncologist's workflow and uses pharma-sponsor subsidies to pay for historically unreimbursed trial-coordination work. Looser eligibility directly raises the match rate at the point of care — more eligible patients per screen = more value per screen and a more rational sponsor subsidy = a stickier product. Candid nuance the team should hear: broader eligibility does NOT fix the reimbursement gap (why community sites don't run trials isn't only that patients get screened out — it's that the coordination work doesn't pay; that unmet need IS the model and this guidance doesn't touch it). And the genuine watch-item: Operation TrialBlazer explicitly includes the HHS OIG and a payment/access-reform workstream — the exact policy space where the anti-kickback/safe-harbor rules around who can pay a physician for trial-related work could be clarified (helpfully) or tightened. Rose Wang's team should read the TrialBlazer payment track as closely as the eligibility guidances — the same initiative that expands the addressable market also governs the legality of the revenue mechanism; make the compliance story airtight before the national rollout. Competitive note: the tailwind isn't proprietary — Massive Bio (FOMAT community-site network, ~July 20) and Tempus (TIME network, phase-1 activation) ride the same thesis; differentiation must stay on the clinician-workflow-plus-subsidy model as raw AI matching commoditizes. https://www.raps.org/resource/fda-finalizes-three-guidances-to-broaden-cancer-clinical-trial-eligibility.html ; https://www.federalregister.gov/documents/2026/07/28/2026-15185/cancer-clinical-trial-eligibility-criteria-performance-status-guidance-for-industry-institutional ; https://www.federalregister.gov/documents/2026/07/28/2026-15187/cancer-clinical-trial-eligibility-criteria-laboratory-values-guidance-for-industry-institutional ; https://www.federalregister.gov/documents/2026/07/28/2026-15186/cancer-clinical-trial-eligibility-criteria-washout-periods-and-concomitant-medications-guidance-for ; https://www.thepharmaletter.com/final-fda-guidances-to-expand-participation-in-cancer-clinical-trials **2. aSKY Therapeutics — no change since 2026-07-26.** No new precision-NSCLC/RAS catalyst since the pair of July 22 FDA actions (GSK's zidesamtinib approved in ROS1+ NSCLC; Revolution Medicines' daraxonrasib NDA accepted in previously-treated metastatic PDAC; neladalkib/ALK still under FDA review). The competitive map stands; the refresh remains on Eddie's list. **3. GPR52 / schizophrenia NewCo (Lieber Institute + Third Rock) — no change since 2026-07-27.** Nxera's Phase-2-ready GPR52 agonist remains unsold this window, so the modest timing reprieve holds; a Q3 pickup by Otsuka / BMS (via Karuna) / Neurocrine / AbbVie (via Cerevel) would reset the differentiation clock. **4. RNA translation-activator platform — no change since 2026-07-28.** No movement beyond yesterday's CAMP4 CMP-002 first-in-human clearance, treated as validation of the turn-the-protein-up modality and a pace-setter (not a threat to a specific asset). Action stands: Avi + platform leads define the defensible wedge.

Portfolio Watch — the RNA-upregulation lane just cleared its first regulatory bar into humans, and it lands on one of our platforms: CAMP4's CMP-002, an antisense oligonucleotide that turns a gene UP (targets regulatory RNA to raise SYNGAP1 protein in haploinsufficient SYNGAP1-related disorder), got Australian TGA/HREC clearance to start a Phase 1/2 first-in-human trial (July 27) — triggering the second close of a financing worth up to $50M (Janus Henderson, Vivo, 5AM, Balyasny, CURE SYNGAP1). Direct read-across to Blackbird's RNA translation-activator platform: validation of the "turn-the-protein-up" modality AND a pace-setter clock on our differentiation. Plus no-change notes on the GPR52 NewCo (Nxera's rival still unsold) and aSKY (RAS map unchanged).

Tue, 28 Jul 2026 13:00:00 +0000

Portfolio Watch for Tuesday, July 28, 2026. One substantive item on an otherwise quiet portfolio-catalyst day — and it sits directly on top of a Blackbird platform. **1. LEAD — CAMP4 Therapeutics' CMP-002 cleared for first-in-human trial (July 27, 2026): direct validation of the RNA translation-activator / upregulation modality.** Australia's Therapeutic Goods Administration (TGA) and local HREC cleared CAMP4 to initiate a Phase 1/2 first-in-human trial of CMP-002, an antisense oligonucleotide (ASO) that works in the OPPOSITE direction from classic knockdown ASOs: it targets regulatory RNA (regRNA) upstream of a gene to release the brake and increase target-protein expression. Target: SYNGAP1. Indication: SYNGAP1-related disorder — a rare haploinsufficient neurodevelopmental condition (>10,000 U.S. patients; universal intellectual disability, epilepsy in ~85%, no approved disease-modifying therapy). Rationale: if disease stems from loss of one working copy, dialing the good copy back toward wild-type levels should be disease-modifying. Preclinical package: dose-dependent SYNGAP protein increases in patient-derived neurons, phenotype reversal in humanized haploinsufficient mice, seizure reduction in a chemically-induced seizure model, broad brain distribution and protein upregulation in non-human primates; intrathecal dosing. Financing: the clearance satisfied conditions for the second close of a September 2025 Securities Purchase Agreement, unlocking up to $50M (investors include Janus Henderson, Vivo Capital, 5AM Ventures, Balyasny, Adage, Coastlands, Trails Edge, and patient foundation CURE SYNGAP1). CAMP4's RAP Platform maps regRNAs and generates ASO candidates across >1,200 haploinsufficient / partial-loss-of-function disorders where a modest protein increase may benefit. **So-what for Blackbird:** this is the clearest validation to date of the "turn-the-protein-up" modality our RNA translation-activator platform lives in — knockdown drugs don't apply to loss-of-function/haploinsufficiency, and CAMP4 (the name to watch in this lane) has now shown a regulator will let an RNA-upregulation drug into humans AND framed the >1,200-disorder commercial prize we should be framing for our own platform. Candid read cuts both ways: validation (not a threat to a specific asset — we have no competing SYNGAP1 program), but a pace-setter — the leading independent RNA-upregulation company is now first-in-human and freshly funded, so the window to establish a differentiated wedge (targeting mechanism vs. regRNA-antisense; delivery route vs. intrathecal; disease set) is a live clock. Action: Avi + platform leads — define the defensible angle now, while the lane is still forming. https://www.globenewswire.com/news-release/2026/07/27/3333378/0/en/CAMP4-Therapeutics-Secures-Australian-Regulatory-Clearance-to-Initiate-First-in-Human-Clinical-Trial-of-CMP-002-in-Patients-With-SYNGAP1-Related-Disorder.html ; https://www.camp4tx.com/science/regrna/ **2. GPR52 / schizophrenia NewCo (Lieber Institute + Third Rock) — no change since 2026-07-27.** Nxera's Phase-2-ready GPR52 agonist (the only other Phase-2-ready GPR52 program; Boehringer walked from its option Dec 2025) remains unsold this window — the modest timing reprieve flagged 07-12/07-27 holds; our NewCo isn't yet racing a resourced same-mechanism Phase 2 rival. Live watch: a Q3 pickup by Otsuka / BMS (via Karuna) / Neurocrine / AbbVie (via Cerevel) would reset the clock. No action today. **3. aSKY Therapeutics — no change since 2026-07-26.** No new precision-NSCLC/RAS catalyst; the competitive map stands (Revolution Medicines' daraxonrasib NDA accepted in PDAC; GSK's zidesamtinib approved in ROS1+ NSCLC; neladalkib/ALK still under FDA review). Nothing to act on beyond the competitive-map refresh already on Eddie's list.

Portfolio Watch — quiet-catalyst week (no Blackbird-company news): a status read on the two fronts that are moving. (1) The GPR52 differentiation clock — Nxera's Phase-2-ready GPR52 agonist NXE'149 remains unsold since Boehringer's December walk-away (a modest reprieve on timing for the Lieber/Third Rock GPR52 NewCo), while the broader non-dopaminergic wave (Cobenfy's gradual launch + missed adjunctive study; Neurocrine's M4 agonist in Phase 3) proves novelty isn't the moat — the DC-declaration value case must be best-in-class chemistry across all three symptom domains. (2) A genetic-medicine tailwind — Arrowhead's GalNAc-siRNA plozasiran (Redemplo) hit both Phase 3 sHTG trials (July 22: TG down up to 81%, acute pancreatitis down 78% / 100% in highest-risk), directly de-risking the quarterly liver-directed siRNA route the hLMR1 (Ruan/JHU) sourcing thesis rests on. Plus an aSKY no-change note.

Mon, 27 Jul 2026 13:00:00 +0000

Portfolio Watch for Monday, July 27, 2026. Honest headline: a quiet week for hard portfolio catalysts — no Blackbird portfolio company printed news this window. Today is a status read on the two threads genuinely in motion. **1. The GPR52 / schizophrenia competitive picture is getting more legible (Lieber Institute + Third Rock GPR52 NewCo; DC declaration expected later 2026).** The direct rival is still Nxera's NXE'149, the only other Phase-2-ready GPR52 agonist. Boehringer walked away from its option in December 2025; Nxera has been out-licensing the program since, targeting a big-pharma or specialty-neuro buyer in 2026 — and as of today no buyer has surfaced. Read: the same-mechanism rival remains unfunded/unpartnered = a modest reprieve on the timing pressure flagged 2026-07-12; our NewCo isn't yet racing a resourced Phase 2 competitor. Live watch: a Q3 pickup by Otsuka, Bristol Myers (via Karuna), Neurocrine, or AbbVie (via Cerevel) resets the clock and puts a same-target rival 12–24 months ahead. The harder signal from the broader non-dopaminergic wave — novelty doesn't auto-convert: Cobenfy (BMS; first muscarinic/non-dopaminergic antipsychotic) has had a gradual launch and its adjunct-to-atypicals schizophrenia study missed; Neurocrine's M4-selective agonist NBI-1117568 is in Phase 3. So-what: at DC declaration the value case can't lean on first-in-class-ness — it must be best-in-class GPR52 chemistry credibly hitting all three symptom domains (positive/negative/cognitive) with a biomarker/PD package, and per Martinowich's 2026-07-10 framing, pair the neuronal-engagement readout with a multicellular tissue-state readout. Action stands: Hemaka + Third Rock — differentiation is the whole ballgame; the market bar is head-to-head-grade proof, not a new receptor. https://www.globenewswire.com/news-release/2025/12/18/3208254/0/en/Nxera-Pharma-to-Regain-Full-Rights-to-GPR52-Agonist-Program-for-Schizophrenia.html ; https://www.fiercebiotech.com/biotech/boehringer-walks-away-nxeras-phase-2-ready-schizophrenia-program ; https://finance.yahoo.com/sectors/healthcare/articles/does-cobenfy-potential-become-top-161300124.html ; https://www.prnewswire.com/news-releases/neurocrine-biosciences-initiates-phase-3-registrational-program-for-nbi-1117568-as-potential-treatment-for-adults-with-schizophrenia-302441930.html **2. Genetic-medicine modality tailwind — Arrowhead plozasiran (Redemplo) Phase 3 win (July 22, 2026).** Topline from Phase 3 SHASTA-3 and SHASTA-4 in severe hypertriglyceridemia: both met all primary/secondary endpoints — triglycerides down up to 81%, acute pancreatitis events down 78% overall and 100% in the highest-risk sHTG patients, favorable safety, once-every-3-months subcutaneous dosing. FDA supplemental filing planned before year-end; full data at ESC Congress Aug 30. Plozasiran is a GalNAc-conjugated siRNA suppressing hepatic APOC3 — the liver-directed RNAi playbook. Not a direct portfolio hit, but it de-risks the modality two Blackbird threads live in (Aletira cell-selective genetic medicines; the RNA-upregulation/translation-activator platform), and it is direct validation of the once-quarterly liver-directed GalNAc-siRNA route the hLMR1 sourcing thesis (Ruan lab, JHU Endocrinology; ranked in the 2026-07-26 weekly radar) is predicated on. Tailwind, not threat. https://ir.arrowheadpharma.com/news-releases/news-release-details/arrowhead-pharmaceuticals-reports-topline-results-phase-3-shasta ; https://www.businesswire.com/news/home/20260722499878/en/Arrowhead-Pharmaceuticals-Reports-Topline-Results-from-Phase-3-SHASTA-3-and-SHASTA-4-Studies-of-Plozasiran-in-Patients-with-Severe-Hypertriglyceridemia **3. aSKY Therapeutics — no change since Sunday.** No new catalyst since the 2026-07-26 update; the map stands (Revolution Medicines' daraxonrasib NDA accepted in PDAC; GSK's zidesamtinib approved in ROS1+ NSCLC; neladalkib/ALK still under FDA review). Nothing to act on today beyond the competitive-map refresh already on Eddie's list.

Sourcing Radar — inaugural weekly roundup: the week's top three deal-ready partner-institution leads — Suk lab UMB aCD47-CpG immune-stimulating antibody conjugate (antigen-agnostic, Fc-independent CD47×TLR9 platform; live patent + TEDCO/MII funding = open UM Ventures window) + Berger/Yegnasubramanian JHU DCTPP1 inhibitors as a decitabine resistance-breaker (provisional filed) + Ruan lab JHU hLMR1 liver-lncRNA MASH/urea-cycle target (no patents disclosed = time-sensitive)

Sun, 26 Jul 2026 12:00:00 +0000

Sourcing Radar for Sunday, July 26, 2026 — the first of the new weekly editions, run as a roundup of the week's three best commercializable leads out of Johns Hopkins / UMB / Lieber, ranked by deal-readiness. A full bioRxiv/medRxiv corresponding-author sweep of 2026-07-20→26 plus a PubMed journals pass confirmed no NEW full-text-eligible, company-shaped partner-institution lead beyond what the week's daily radars already surfaced (three fresh company-shaped preprints remain abstract-only deferrals; see closing note), so today consolidates the standouts. **Lead (full text read on 2026-07-25) — Suk lab UMB aCD47-CpG immune-stimulating antibody conjugate (ISAC).** Kong (first), Suk senior corresponding (UMB Neurosurgery + UM-MIND; jsuk@som.umaryland.edu), "An immune-stimulating antibody conjugate spatiotemporally targeting CD47 and TLR9 elicits macrophage-dependent tumor clearance and durable anti-cancer adaptive immunity," bioRxiv v1 2026-07-23 (DOI 10.64898/2026.07.22.739359). Covalently links an anti-CD47 antibody to a TLR9-agonist CpG oligo (DAR 3); because TLR9 is endosome-restricted, the same macrophage that engulfs the tumor (CD47 blockade) also gets the intracellular co-stim signal — converting immunologically silent phagocytosis into M1 repolarization, antigen cross-presentation, CD8+ priming, and durable memory. In vivo: macrophage-dependent clearance of human NHL xenograft (abolished by clodronate); extended survival + day-106 rechallenge resistance in immunocompetent NHL; tumor + lung-metastasis suppression and Treg depletion in cold 4T1 TNBC. COI verbatim: "J.S.S. and S.W.C. are inventors on a patent application related to the aCD47-CpG conjugate technology…"; funded by NIH R01NS119609 + Maryland Innovation Initiative/TEDCO. So-what: unlike antigen-restricted, FcγR-dependent clinical ISACs (Bolt BDC-1001 et al.), this rides a ubiquitous pan-cancer myeloid checkpoint and works Fc-independently (Fc-silent backbone possible) = differentiated platform; live patent + state funding = open UM Ventures window now; Suk moved from the JHU Center for Nanomedicine cluster to UMB's UM-MIND = UM Ventures × Blackbird co-investment surface. Actions: Eddie — filing status + is Suk contemplating a vehicle + BioHub/BioVentures fit; Avi/Yixuan — FTO vs the anti-CD47 antibody estate + ISAC linker patents. Strongest UMB lead of the month. **Pick 2 (flag-and-chase — PNAS body paywalled, briefed from Significance+abstract+JHU press) — Berger/Yegnasubramanian/Nelson JHU DCTPP1 inhibitors as a decitabine resistance-breaker.** Hauk (first), Berger + Yegnasubramanian co-corresponding, "Structural and cellular insights into DCTPP1 antagonists and their synergistic action with DNMT inhibitors," PNAS 123(25):e2534029123, 2026-06-15 (DOI 10.1073/pnas.2534029123; PMID 42296362). DCTPP1 is a nucleotide-pool-surveilling pyrophosphatase that clears modified deoxynucleotides; since nucleoside-analog DNMT inhibitors (decitabine, azacitidine) work by DNA misincorporation, DCTPP1 is a cell-intrinsic resistance factor — inhibit it to potentiate an already-approved generic. 10,000-compound HTS → sub-micromolar inhibitors; crystal structures of three chemical classes in the pocket; synergy with decitabine in prostate-cancer cells. So-what: NCE (the inhibitor) on a cheap generic backbone in an existing market; provisional patent filed = open JHTV window; heavyweight founding nucleus (Kimmel director Nelson, IBBS director Berger, inHealth's Yegnasubramanian). Diligence: chemical-matter novelty + FTO (they describe both new and known inhibitor classes). **Pick 3 (full text read, open access; most time-sensitive on IP) — Ruan lab JHU hLMR1 hepatocyte-specific liver-lncRNA metabolic target.** Jaso-Vera (first), Ruan senior corresponding (JHU Endocrinology), "hLMR1, a hepatocyte-specific long noncoding RNA that represses amino acid catabolism through pre-mRNA interaction in human liver," PLoS One 21(7):e0353674, 2026-07-21 (DOI 10.1371/journal.pone.0353674; PMID 42479790). hLMR1 is a liver-specific AND human-specific lncRNA that brakes hepatic amino-acid catabolism + ureagenesis via a 12-nt pre-mRNA-hybridization motif; knockdown in humanized-liver mice upregulates AA-degradation + urea-cycle genes. So-what: first-in-class liver-restricted repressor RNA — you win by knocking it down; authors nominate ASO or GalNAc-siRNA = the de-risked Ionis/Alnylam liver playbook; MASH field all works lipid/fibrosis, nobody drugs hepatic nitrogen metabolism, and it opens a urea-cycle-disorder orphan lane. Early (target validation, no molecule) = right venture-studio stage. COI: no patents disclosed = JHTV filing window likely still open — TIME-SENSITIVE. Action: Eddie — Ruan + JHTV filing-status conversation. **Still cooking (abstract-only deferrals, full text not yet rendered — re-read next week):** Kwon lab JHU iPSC-cardiomyocyte in-vivo bioincubation maturation platform (DOI 10.64898/2026.07.21.739858); Krantz lab UMB single-molecule-proteomics-via-dynamical-translocases perspective (DOI 10.64898/2026.07.20.739629); Watanabe/Bergles JHU "skoupocytosis" microglial neuronal-organelle clearance, druggable ABHD16a (DOI 10.64898/2026.07.22.740182). Paper links: https://www.biorxiv.org/content/10.64898/2026.07.22.739359v1 ; https://doi.org/10.1073/pnas.2534029123 ; https://doi.org/10.1371/journal.pone.0353674

Portfolio Watch — no portfolio-company news this window, but the precision-NSCLC / RAS landscape hardened around aSKY Therapeutics in a single day: FDA approved GSK/Nuvalent's zidesamtinib (Jideytro) in ROS1+ NSCLC (July 22, GSK's first-ever lung approval) and accepted Revolution Medicines' daraxonrasib RAS(ON) NDA in pancreatic cancer (July 22) — validation + a rising differentiation bar for aSKY's undisclosed KRAS-frame mechanism; plus a 1104health competitor note (Massive Bio site-network expansion)

Sun, 26 Jul 2026 13:00:00 +0000

Portfolio Watch for Sunday, July 26, 2026. Honest headline: no Blackbird portfolio company printed news in the past ~4 days. The live signal is the competitive ground moving under aSKY Therapeutics (UMB/UM Ventures × Blackbird oncology spinout; KRAS-frame NSCLC, MOA undisclosed) — two regulatory milestones landed the same day. **1. FDA approves GSK's zidesamtinib (Jideytro) in previously-treated ROS1+ NSCLC — July 22, 2026.** A ROS1-selective TKI, cleared ~2 months ahead of its PDUFA date; GSK's first-ever lung-cancer approval and a fast payoff on the ~$10.6B Nuvalent acquisition. Sibling neladalkib (ALK) NDA remains under FDA review (Nov 27 target). aSKY hook: zidesamtinib + neladalkib are exactly the GSK-Nuvalent competitive axis flagged in aSKY's watch. Read: precision-oncology dealmaking in lung is paying off on a compressed timeline (category validation), but each new selective approval raises the bar for aSKY's still-undisclosed mechanism to define a defensible lane. Caveat: different drivers (ROS1/ALK, not KRAS) = landscape validation more than head-to-head threat. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-zidesamtinib-ros1-positive-non-small-cell-lung-cancer **2. FDA accepts Revolution Medicines' daraxonrasib NDA in previously-treated metastatic pancreatic cancer — July 22, 2026.** Oral multi-selective RAS(ON) inhibitor; NDA accepted on Phase 3 RASolute 302 data and pulled into the FDA Commissioner's National Priority Voucher pilot. PDAC is the lead indication; the broader Phase 3 program includes NSCLC. aSKY hook: the more direct KRAS/RAS comp — Revolution is the lead RAS(ON) player, and a KRAS-frame lung program must define what it does that daraxonrasib and the covalent-KRAS crowd don't. Read: RAS-mutant oncology is de-risking fast at the regulatory level = tailwind for the science AND a differentiation pressure-test. https://www.globenewswire.com/news-release/2026/07/22/3331714/0/en/revolution-medicines-new-drug-application-for-daraxonrasib-accepted-for-review-by-u-s-fda-for-previously-treated-metastatic-pancreatic-cancer.html **Action:** diligence hygiene, not a fire drill — Eddie + the aSKY team should refresh the competitive map now that the KRAS/RAS field has two more validating datapoints and a clearer sense of who reaches the FDA first; the undisclosed-mechanism gap gets more expensive as the field matures. **Landscape note (just outside the window, ~July 20) — 1104health competitor activity.** Massive Bio (AI patient-matching) expanded its cancer-trial prescreening onto a large clinical-site network (FOMAT). Trend to watch: consolidation + AI matching hardening into incumbents; doesn't change 1104health's clinician-workflow + pharma-subsidy thesis, but it's the competitive weather. https://www.morningstar.com/news/business-wire/20260720835336/massive-bio-partners-with-fomat-to-expand-cancer-patient-prescreening-for-clinical-trial-access

Sourcing Radar — Fifth thin JHU/UMB/Lieber preprint week, but a strong full-text-read UMB lead: Suk lab (UMB Neurosurgery / UM-MIND) aCD47-CpG immune-stimulating antibody conjugate spatiotemporally couples CD47 blockade to endosomal TLR9 (CpG) stimulation, converting silent phagocytosis into durable anti-tumor immunity across NHL + TNBC (bioRxiv 2026-07-23; live patent filing; TEDCO/Maryland Innovation Initiative + NIH funded = open UM Ventures window) + Ferrer lab UMB Greenebaum GSTT1/CD133/FGFR pancreatic-cancer stem-like biomarker-repurposing play (Cancer Letters, abstract-only) + Watanabe/Bergles JHU "skoupocytosis" microglial neuronal-organelle clearance via ABHD16a (target-biology watch, body not yet rendered)

Sat, 25 Jul 2026 12:00:00 +0000

Sourcing Radar for Saturday, July 25, 2026. Fifth consecutive thin JHU/UMB/Lieber preprint week: the four-day bioRxiv/medRxiv corresponding-author sweep (2026-07-21→25; 1092 bioRxiv + 262 medRxiv records via api.biorxiv.org details API, filtered on author_corresponding_institution) surfaced only a handful of partner-institution preprints, and all fresh company-shaped ones except the lead had unrendered bodies (deferrals). Peer-reviewed journals in-window were again mostly clinical trials, reviews, and collaborator-only work (the Cancer Immunol Res CD44 stromal-checkpoint PDAC paper is Georgetown-corresponding with JHU/UMB co-authors; the Dev Cell KRAS-mutant PDAC paper is Weill Cornell-corresponding). One item was full-text-readable and clears the bar — and it is the strongest UMB lead of the month. **LEAD (FULL TEXT READ via PDF) — Suk lab UMB aCD47-CpG immune-stimulating antibody conjugate (ISAC).** Byoungjae Kong (first), Seung Woo Chung (JHU Ophthalmology), Daiheon Lee, Gijung Kwak, Jung Soo Suk (senior corresponding; jsuk@som.umaryland.edu), Department of Neurosurgery + University of Maryland Medicine Institute for Neuroscience Discovery (UM-MIND), University of Maryland School of Medicine (Suk also holds JHU Ophthalmology/Neurosurgery + UMD College Park Fischell Bioengineering appointments). "An immune-stimulating antibody conjugate spatiotemporally targeting CD47 and TLR9 elicits macrophage-dependent tumor clearance and durable anti-cancer adaptive immunity," *bioRxiv* v1 2026-07-23; DOI 10.64898/2026.07.22.739359. **COI verbatim: "J.S.S. and S.W.C. are inventors on a patent application related to the aCD47-CpG conjugate technology described in this manuscript. The remaining authors declare no competing interests."** Funding: NIH R01NS119609 (J.S.S.); **Maryland Innovation Initiative (MII) of TEDCO, 0722-005 (J.S.S.)**; Sejong Science Fellowship, NRF Korea (B.K.); Greenebaum CCC flow core. **Science.** CD47 is the "don't-eat-me" macrophage checkpoint; anti-CD47 antibodies (magrolimab/Gilead, ALX Oncology, Pfizer/Trillium) drive phagocytosis but clearance is "immunologically silent" — no durable adaptive immunity, tumors recur. aCD47-CpG covalently links an anti-CD47 antibody (mouse clone MIAP410; human clone B6H12) to a TLR9-agonist CpG oligo via non-cleavable SMCC crosslinker, CpG conjugated at the 3′-end to preserve potency; DAR 1–4 screened, DAR 3 optimal (NF-κB plateaus beyond DAR 3). Rationale for TLR9 specifically: it is restricted to the endosome/phagolysosome — the same compartment where engulfed tumor antigen is processed — so the conjugate guarantees the SAME macrophage that engulfs the tumor (via CD47 blockade) receives the intracellular co-stimulatory signal, a spatiotemporal coupling unattainable by co-dosing two free drugs. Results: retains nanomolar CD47 + TLR9 binding; conjugate shows an initial phagocytosis delay (attenuated CD47 affinity) but surpasses naked aCD47 after ~8 h (TLR9-driven sustained phagocytosis); upregulates CD80/CD86/CD40 + iNOS (M1) and TNFα/IL-6/IL-12; boosts MHC-I antigen cross-presentation and OT-I CD8+ T-cell proliferation + IFN-γ (naked aCD47 does not). In vivo: MTD 10 mg/kg (transient reversible effects, no priming needed); tumor-dominant biodistribution despite CD47 antigen sinks; human Raji NHL xenograft — profound regression, abolished by clodronate macrophage depletion (macrophage-dependent), M1:M2 shift only with conjugate; immunocompetent systemic A20 NHL — reduced burden, extended survival, complete resistance to day-106 rechallenge (durable memory); orthotopic syngeneic 4T1 TNBC (cold, immunosuppressive) — tumor suppression, enriched macrophages/monocytes/CD8+/effector + central-memory T cells, Treg depletion, near-complete abrogation of lung metastases; human MDA-MB-231 TNBC recapitulated key responses. **So what for Blackbird.** (1) Differentiation: every clinical ISAC (Bolt Biotherapeutics BDC-1001 and cohort) hangs a TLR7/8/9 or STING payload off a tumor-specific antibody (HER2, EGFR, CD22) and needs FcγR engagement — antigen-restricted. aCD47-CpG exploits a ubiquitous pan-cancer myeloid checkpoint, lowers the phagocytic threshold intrinsically, and works Fc-independently (potent on a mouse IgG1 backbone with minimal activating-FcγR affinity), enabling an Fc-silent backbone to mitigate systemic toxicity — antigen-agnostic + Fc-independent = a differentiated platform position (contrast the recent aCD47-listeriolysin-O conjugate, which needs cytosolic release). (2) Dealability: live patent application (Suk + Chung inventors) + TEDCO/MII state translational funding = an open UM Ventures licensing window right now. (3) Ecosystem: Suk came up through the JHU Center for Nanomedicine cluster and is now building this out of UMB's new UM-MIND institute — a serious nanomedicine PI on the UM Ventures × Blackbird co-investment surface alongside the tracked Chao (TLR7-ITGAM) and Martin (MCAM/TetherChip) programs. **Actions:** Eddie — UM Ventures filing status + whether Suk is contemplating a vehicle; is Blackbird BioHub/BioVentures a fit at this stage. Avi/Yixuan — freedom-to-operate against the anti-CD47 antibody estate and ISAC linker patents (conjugate chemistry is where value + risk sit); Fc-silencing + human-conjugate manufacturability. Added Jung Soo Suk to tracked-PI list and a new UMB aCD47-CpG ISAC entry; strongest UMB lead of the month. **Pick 2 (ABSTRACT-ONLY — Cancer Letters is Elsevier-paywalled, NOT a full read; flag-and-chase) — Ferrer lab UMB Greenebaum GSTT1/CD133/FGFR pancreatic-cancer stem-like axis.** Deborah de la Caridad Delgado Herrera (first) … Elana J. Fertig (co-author; already tracked), Christina M. Ferrer (senior corresponding; cferrer@som.umaryland.edu), University of Maryland School of Medicine + Marlene and Stewart Greenebaum Comprehensive Cancer Center + UMB Institute for Genome Sciences. "GSTT1 promotes stemness and FGFR inhibitor sensitivity in pancreatic cancer through regulation of CD133 (PROM1)," *Cancer Letters* 2026 (in press); DOI 10.1016/j.canlet.2026.218733. **Science (abstract):** a slow-cycling, highly metastatic GSTT1+ PDAC subpopulation is stem-like (tumor-sphere formation, PROM1/CD133, Wnt + FGF signaling); CD133+GSTT1+ human PDAC cells define a maximal stem-like state; FGFR-inhibitor sensitivity is context-dependent (only under tumor-sphere conditions); FGFR3 correlates with GSTT1/CD133; GSTT1 knockdown lowers CD133 and alters FGFR-inhibitor sensitivity; in patient-derived organoids, GSTT1/PROM1 co-expression predicts enhanced response to the approved multi-kinase inhibitor nintedanib. **So what:** GSTT1 is germline-polymorphic (common whole-gene null) — a ready-made patient-stratification biomarker; the commercial shape is a companion-biomarker + repurposed-approved-drug play in PDAC on the UM Ventures × Blackbird co-investment surface, not a new molecule. Diligence: whether Greenebaum has filed on the biomarker; lower conviction than the Suk program until the full paper is readable. Added Christina Ferrer + GSTT1/CD133/FGFR-in-PDAC to tracked list. **Watch (ABSTRACT-ONLY — body not yet rendered, DEFERRAL to re-read) — Watanabe/Bergles JHU "skoupocytosis."** Rachel Pepper (first) … Dwight E. Bergles, Shigeki Watanabe (senior corresponding), Johns Hopkins University (Cell Biology / Neuroscience). "Microglia extract neuronal proteolytic organelles via skoupocytosis," *bioRxiv* v1 2026-07-24; DOI 10.64898/2026.07.22.740182. **Science (abstract):** microglia make transient contact with neuronal membranes and pinch off a small portion of the process containing an oversized proteolytic organelle (leaving the rest intact) — a newly named clearance route ("skoupocytosis," after the Greek for garbage) that bypasses retrograde organelle transport; the PS lipase ABHD16a accumulates at these sites and converts phosphatidylserine to lyso-PS to initiate the process. **So what:** no asset, no disclosed IP, and the body text has not rendered yet (couldn't read in full) — but a defined, druggable enzyme (ABHD16a) gating neuronal proteostasis is seed-grade neurodegeneration target biology. Target-biology watch + deferral; added Shigeki Watanabe + ABHD16a/skoupocytosis to tracked PI/target list, re-read on full-text render. Paper links: https://www.biorxiv.org/content/10.64898/2026.07.22.739359v1 ; https://doi.org/10.1016/j.canlet.2026.218733 ; https://www.biorxiv.org/content/10.64898/2026.07.22.740182v1

Sourcing Radar — Fourth thin JHU/UMB/Lieber preprint week: fully-read anchor is Green/Tzeng JHU BME PBAE non-viral gene-delivery plasmid-engineering paper (antibiotic-free minimal Nanoplasmid = CMC/manufacturing-readiness win for liver + brain cancer), PLoS One 2026-07-23 — but the platform is already spun out (Green at Cove, Dome, OncoSwitch; Tzeng co-founded OncoSwitch) so this is ecosystem intelligence, not a fresh deal + BACH1 pioneer-repressor macrophage-plasticity target biology from Nagy lab JHU (Immunity 2026-07-17, abstract-only/paywalled = target-watch + relationship). No item cleared the founding bar this week

Fri, 24 Jul 2026 12:00:00 +0000

Sourcing Radar for Friday, July 24, 2026. A fourth consecutive thin JHU/UMB/Lieber week: the four-day bioRxiv/medRxiv corresponding-author sweep (2026-07-20→24; 1072 bioRxiv + 259 medRxiv records via api.biorxiv.org details API, filtered on author_corresponding_institution) returned NO partner-institution preprint with a new ownable asset — only items already logged (Kim C. elegans FOXO/DAF-16 dauer dev-bio; Kwon iPSC-cardiomyocyte in-vivo bioincubation, body still not rendered; Yi zebrafish swim-bout NN methods; Chanda PCI+TAVR renal-outcomes meta-analysis; Ranjan PNH complement-vs-inflammasome [passed 07-22]; Higgins-Tejera social-adversity aging clocks). Peer-reviewed journals in-window were heavy on clinical trials/reviews/collaborator-only work (Azad JHU copanlisib+nivolumab MSS-CRC negative Ph1/2 Nat Commun; Cell autoantibody-immunosurveillance commentary [JHU Dang co-author, CDI Labs Baltimore co-author]; Pitt motor-thalamus DBS speech/swallowing [JHU co-author]; Henan-led EnzGFM enzyme PLM [JHU minor co-author, already logged 07-18]; UMD-College-Park amorphous ferroelectric = out of scope). The Oct-2025 JHU/UMB osteosarcoma CGRP/TrkA/NGF pain-drug repurposing PNAS paper (Ramesh/James) re-surfaced via a fresh EurekAlert push but is out of the recency window. So two watch-grade Johns Hopkins items, and NEITHER is a clean spin-out. **Anchor (FULL TEXT READ, open access) — Green/Tzeng JHU non-viral gene-delivery plasmid-engineering paper.** Jordan J. Green (senior) + Stephany Y. Tzeng, Department of Biomedical Engineering, Johns Hopkins University School of Medicine (+ Oncology/Sidney Kimmel CCC, INBT, Bloomberg~Kimmel, Materials Science, Neurosurgery, Ophthalmology). "Engineering of episomal plasmid structure to enhance non-viral Poly(beta-amino ester) nanoparticle gene delivery to liver and brain cancer cells," *PLoS One* 2026-07-23; DOI 10.1371/journal.pone.0352468; PMID (edat 2026-07-23). **COI verbatim: Johns Hopkins has filed patents on the polymer technology with co-inventors Green and Tzeng; Green holds positions at Cove Therapeutics, Dome Therapeutics, and OncoSwitch Therapeutics; Tzeng is manager/co-founder of OncoSwitch Therapeutics.** Funding: NIH R01CA228133, R37CA246699, P41EB024495, R01EY031097, P41EB028239 + Goldhirsh-Yellin Foundation. **Science.** 14 episomal reporter plasmids varying promoter (CMV, CAG, EF1α), backbone (Z1, pUNO1, Nanoplasmid), length (~2000–7000 bp) and antibiotic marker (Kanamycin/Zeocin/Blasticidin/none), packaged in PBAE nanoparticles and tested across 6 HCC lines (human Hep3B; murine Hepa1-6, Hepa1c1c7; porcine A92, A272, B239) and 3 brain-cancer lines (murine glioma CT-2A, human astrocytoma CCF-STTG1, human meningioma IOMM-Lee). CMV out-performed CAG/EF1α at day 3; the minimal Nanoplasmid (~2,096 bp, GFP) gave the highest %GFP+ in 5/6 HCC lines and highest gMFI in all 6, plus best luciferase in 2/3 brain lines — and is the ONLY construct without an antibiotic-resistance gene, an advantage for FDA clinical-translation guidance. Modest negative size-vs-efficacy correlation. **So what for Blackbird (candid).** Strong platform, but incremental optimization of an already-spun-out franchise — value largely captured (Cove/Dome/OncoSwitch). Real Blackbird value = ecosystem intelligence on a Baltimore gene-therapy company cluster from one JHU lab, adjacent to the tracked Mao-lab anionic-LNP delivery chassis; plus the antibiotic-free minimal-plasmid CMC detail as a de-risking data point for any non-viral program. NOT a fresh sourcing opportunity. **Action:** Eddie — map OncoSwitch/Dome/Cove financing + indication split. Added Jordan Green + Stephany Tzeng to tracked-PI list; Cove/Dome/OncoSwitch to tracked-company list. **Watch (ABSTRACT-ONLY — Immunity is Elsevier-paywalled, NOT a full read) — BACH1 as a "pioneer repressor" of macrophage plasticity.** László Nagy (senior corresponding; lnagy@jhmi.edu), Department of Medicine, Johns Hopkins University School of Medicine / Institute for Fundamental Biomedical Research at Johns Hopkins All Children's Hospital (St. Petersburg, FL), with Debrecen/Warsaw/Cambridge/Leicester collaborators. "The transcription factor BACH1 couples chromatin priming and repression to enable macrophage plasticity and adaptation," *Immunity* 2026-07-17; DOI 10.1016/j.immuni.2026.06.022; PMID 42468527. **Science (abstract):** heme-regulated repressor BACH1 binds inactive + active regulatory regions incl. latent enhancers, recruits NuRD, and has dual function — establishing early chromatin accessibility while actively repressing transcription ("pioneer repression"); on inflammatory stimulation it redistributes in cis to nearby promoters, constraining 3D chromatin architecture; Bach1 deletion in vivo impairs macrophage polarization, tissue adaptation, and resilience during systemic + regenerative inflammation. **So what:** BACH1 is already an active drug-discovery target elsewhere; this sharpens tuning it for macrophage-driven inflammation/fibrosis/regeneration — but it's fundamental mechanism, no asset, no disclosed IP. Target-biology watch + relationship with a heavyweight immunometabolism PI now inside our primary partner institution. Added László Nagy to tracked-PI list. Paper links: https://doi.org/10.1371/journal.pone.0352468 ; https://doi.org/10.1016/j.immuni.2026.06.022

Sourcing Radar — DCTPP1 as a decitabine resistance-breaker: Berger/Yegnasubramanian/Nelson JHU (IBBS + Kimmel) find sub-micromolar DCTPP1 inhibitors (3 crystal-structure classes) that synergize with the hypomethylating agent decitabine in prostate cancer, PNAS 2026-06-15 (provisional patent filed = live JHTV; flag-and-chase, body paywalled) + Jain lab JHU host-directed TB-meningitis therapy in a new immune-vascularized human brain-organoid model, Nat Commun 2026-07-21 (repurposed generics, no IP = platform-watch on the organoid, not a spinout) — third thin preprint week

Thu, 23 Jul 2026 12:00:00 +0000

Sourcing Radar for Thursday, July 23, 2026. A third consecutive thin JHU/UMB/Lieber preprint week (four-day bioRxiv/medRxiv corresponding-author sweep found no ownable asset — a C. elegans dev-bio paper, a stem-cell cardiomyocyte-maturation preprint whose full text hasn't rendered, and clinical/epi work with no drug hook), so two peer-reviewed Johns Hopkins picks with precise vetting notes. **Pick 1 (LEAD as a company-shaped sourcing flag — NEWSROOM/ABSTRACT-GROUNDED, body paywalled) — DCTPP1 inhibitors as a decitabine resistance-breaker.** Glenn Hauk (first), Jianyong Liu, William G. Nelson (Kimmel Cancer Center director), Srinivasan (Vasan) Yegnasubramanian (co-corresponding; Oncology/Pathology/Radiation Oncology; director, inHealth precision medicine; syegnasu@jhmi.edu), James M. Berger (co-corresponding; director, JHM Institute for Basic Biomedical Sciences; jmberger@jhmi.edu), all Johns Hopkins University School of Medicine + Sidney Kimmel Comprehensive Cancer Center. "Structural and cellular insights into DCTPP1 antagonists and their synergistic action with DNMT inhibitors," *PNAS* 123(25):e2534029123, published 2026-06-15; DOI 10.1073/pnas.2534029123; PMID 42296362. **Vetting note: PNAS body is subscription-gated (not OA, not in PMC/EPMC, no preprint) — briefed from the authors' full Significance section + Abstract + the Johns Hopkins/EurekAlert press release (July 15, 2026), NOT a full read. Treat as a flag-and-chase lead, not finished diligence.** **COI verbatim: "The authors have filed for a provisional patent related to this work."** Funding: NCI R35-CA263778 (Berger), P50-CA272391 prostate SPORE (Yegnasubramanian), P30-CA006973 (Sidney Kimmel CCC); Commonwealth Foundation, Irving Hansen Foundation, Prostate Cancer Foundation (Yegnasubramanian). PDB 9YSM–9YSP. **Science (per Significance + Abstract).** DCTPP1 is a nucleotide pyrophosphatase that surveils the dNTP pool and hydrolyzes deoxy-5-methylcytosine triphosphates and other modified nucleotides to prevent their misincorporation into DNA, preserving genomic stability and epigenetic programming. Because nucleoside-analog DNMT inhibitors (decitabine; azacitidine) work by getting misincorporated into DNA, DCTPP1 acts as a cell-intrinsic resistance factor that degrades their efficacy — inhibit DCTPP1 and you potentiate the drug. A 10,000-compound HTS (JHU ChemCore) yielded existing and previously unreported inhibitor classes with sub-micromolar potency; X-ray crystallography solved three distinct inhibitor classes bound in DCTPP1's nucleotide-binding pocket (contacting a tryptophan pair + two critical histidines that mimic substrate); biochemical assays with modified scaffolds confirmed the interactions; cell-based experiments showed significant synergy between lead inhibitors and decitabine in blocking prostate cancer cell growth, with loss- and gain-of-function studies confirming DCTPP1-dependence. **So what for Blackbird.** A resistance-breaker layered on already-approved oncology drugs — installed market in MDS/AML plus a demonstrated path into solid tumors (prostate) — with a novel-target thesis, tool inhibitors already at crystal-structure resolution, and a provisional patent already filed (live JHTV window). Value capture is the new chemical entity (the inhibitor) on a generic backbone. Heavyweight founding/SAB nucleus (Berger structural biology + IBBS; Yegnasubramanian + Nelson cancer/clinical/inHealth). Diligence gaps before it's real: chemical-matter novelty / freedom-to-operate (authors describe both new and known inhibitor classes), and combination-drug regulatory + pricing story. **Actions:** Eddie — JHTV filing status + whether Berger/Yegnasubramanian are contemplating a vehicle. Avi/Yixuan — pull the full paper + patent when it publishes; pressure-test chemistry novelty + FTO. Added to tracked-PI list; Berger already logged (07-21 ORC–HP1 co-author) now senior-corresponding on an oncology druggable-target paper. **Pick 2 (fresh, FULL TEXT READ — platform-watch/relationship, not a spinout) — Jain lab JHU host-directed therapy for TB meningitis in a new human brain-organoid model.** Carlos E. Ruiz-Gonzalez (first), Medha Singh, … William R. Bishai, Sanjay K. Jain (senior corresponding), Johns Hopkins University (Center for Infection and Inflammation Imaging Research). "Host-directed treatments for tuberculous meningitis utilizing a multi-platform approach across mouse and human models," *Nat Commun* 2026-07-21; DOI 10.1038/s41467-026-75202-6; PMID 42481504. **COI verbatim: "The authors declare no competing interests."** Funding: NIH R01-AI145435-A1, R01-AI153349, R56-AI179012-A1, R01-AI190038, S10-OD030381-A1 (all S.K.J.) + NIAID intramural. **Science.** TB meningitis damage is driven by CNS neuroinflammation; steroids (dexamethasone) cut mortality ~25% but not neurological deficits. Twelve immunomodulatory drugs (11 classes; FDA-approved or in trials elsewhere) were tested as adjuncts to isoniazid across a mouse TB-meningitis model (mortality, neuro-deficit, MRI endpoints), a newly developed immune-vascularized human brain organoid model of TB meningitis, and patient PBMCs. Bestatin (aminopeptidase inhibitor), imatinib (TKI), roflumilast (PDE-4 inhibitor), and a PARP-1 inhibitor significantly reduced both mortality and neurological deficits vs the dexamethasone reference, via suppression of neuroinflammation. **So what for Blackbird.** Candid: the drugs are repurposed generics (no COI, no patent) and TB meningitis is a global-health indication (BARDA/foundation economics, not venture returns) — not a spin-out on the drug side. The investable/licensable asset is the newly built immune-vascularized human brain organoid neuroinflammation model — a CNS screening chassis adjacent to the Baltimore imaging/model platforms already tracked (MIPOT, 3DEEP). Plus a relationship worth having (Jain/Bishai infectious-disease + molecular-imaging group; Jain has imaging-tracer IP history). Log as platform-watch + relationship item; look if the organoid model is spun out/licensed as a neuroinflammation screening tool. Paper links: https://doi.org/10.1073/pnas.2534029123 ; https://doi.org/10.1038/s41467-026-75202-6

Sourcing Radar — Ruan lab JHU (Endocrinology + Johns Hopkins All Children's) hLMR1 hepatocyte-specific, human-specific lncRNA that represses hepatic amino-acid catabolism + ureagenesis in MASLD/MASH in PLoS One 2026-07-21 (first-in-class liver-restricted RNA target, oligonucleotide-tractable via ASO/GalNAc-siRNA, no patents disclosed = open JHTV provisional window) + Raabe lab JHU Pediatric Oncology CoREST-inhibitor corin (bifunctional LSD1/HDAC) differentiation therapy extends survival in SMARCB1-null ATRT orthotopic xenografts in Neuro-Oncology 2026-07-21 (rare-pediatric-oncology PRV lane, but corin composition-of-matter is Cole/Harvard = JHU owns use-not-molecule) — dry preprint week, two peer-reviewed JHU picks

Wed, 22 Jul 2026 12:00:00 +0000

Sourcing Radar for Wednesday, July 22, 2026. A thin JHU/UMB/Lieber preprint week (no corresponding-author commercializable preprint in the 2026-07-18→22 bioRxiv/medRxiv window), so two peer-reviewed Johns Hopkins picks that landed this week — one a genuinely novel, ownable target, one a strong translational result with a more complicated IP story. **Pick 1 (LEAD, novel-target sourcing lead, full text read) — Ruan lab JHU hLMR1 liver-lncRNA target in MASLD/MASH.** Jaso-Vera ME (first), Takaoka S, Faccioli LAP, Hu Z, Soto-Gutierrez A, Ruan X (senior corresponding; Division of Endocrinology, Diabetes & Metabolism, Johns Hopkins University School of Medicine + Institute for Fundamental Biomedical Research, Johns Hopkins All Children's Hospital). "hLMR1, a hepatocyte-specific long noncoding RNA that represses amino acid catabolism through pre-mRNA interaction in human liver," *PLoS One* 21(7):e0353674, 2026-07-21. DOI 10.1371/journal.pone.0353674. PMID 42479790. **COI verbatim: "The authors have no relevant financial or non-financial interests to disclose"; no patents mentioned = provisional window likely still open.** Funding: JHU institutional funds + AHA 23SCEFIA1156649 (Ruan); AHA 24DIVSUP1291162 (Jaso-Vera); JHACH ACRI fellowship (Takaoka); NIH DK099257 (Soto-Gutierrez); U. Pittsburgh Center for Transcriptional Medicine internal funds. **Science.** hLMR1 ("human liver metabolic regulator 1") is a hepatocyte-specific, human-specific long noncoding RNA that acts as a brake on hepatic amino-acid catabolism and ureagenesis — the pathways that are impaired, with no prior molecular explanation, in MASLD (formerly NASH). In a humanized-liver mouse (human hepatocytes replacing ~15–80% of the mouse liver by circulating human albumin), knockdown broadly upregulates amino-acid degradation genes (SDS, GLS2, AASS) and urea-cycle enzymes (ASS1, CPS1, ARG1). Mechanism: hLMR1 base-pairs to the pre-mRNAs of those catabolic genes through a 12-nt complementary motif (nucleotides 587–598); full-length overexpression completely blocks glucagon-induced SDS/CPS1/HGD induction, while the Δ587–598 motif-deletion mutant has no inhibitory effect. Human relevance: expression climbs from healthy liver → MASH with reciprocal downregulation of SDS/ASS1; a one-week low-carbohydrate intervention lowers hLMR1 and raises the catabolic enzymes. **So what for Blackbird.** A first-in-class, not me-too, target: a liver-restricted, human-specific repressor RNA — you win by knocking it down, and the authors explicitly nominate antisense oligonucleotides or GalNAc-conjugated siRNA ("pharmacological suppression of hLMR1—using antisense oligonucleotides or GalNAc-conjugated siRNAs—could simultaneously ameliorate hyperlipidemia and hyperglycemia"). That is the most de-risked modality in the business (Ionis/Alnylam GalNAc-to-liver playbook), pouring drug exactly and only into hepatocytes, where this target lives. Competitive backdrop: MASH just got its first approval in resmetirom (Madrigal) and is crowded with FGF21/incretin players — all working lipid/fibrosis biology; nobody is drugging hepatic nitrogen metabolism. The amino-acid/urea-cycle angle also opens a second door toward hyperammonemia / urea-cycle-disorder orphan indications where a liver siRNA that restores ureagenesis is a clean story. IP tell: no COI, no patents mentioned → JHTV provisional window very likely open now. Candid read: early (target validation in humanized mouse, no therapeutic molecule yet) but that is the right stage for a venture studio — found a company around ownable, foundational biology rather than buy a program. Strongest sourcing lead of the week. **Actions.** Eddie — Ruan + JHTV filing-status conversation, time-sensitive given no patents disclosed. Avi/Yixuan — target-novelty + freedom-to-operate diligence; is the 587–598 pre-mRNA-hybridization mechanism defensible and modality-agnostic; human-specificity as both moat and preclinical-model constraint (humanized-liver mice only). Add Xiangbo Ruan (JHU Endocrinology) to tracked-PI list — liver-lncRNA target-discovery lab worth a standing line. **Pick 2 (translational watch / partner item, abstract-grounded — OUP full text CAPTCHA-gated this run) — Raabe lab JHU Pediatric Oncology CoREST-inhibitor corin differentiation therapy in ATRT.** Geethadevi A (first), … Cole PA (Harvard Medical School/Brigham & Women's), Rubens J (now at Pfizer Inc.), … Raabe EH (senior corresponding; Division of Pediatric Oncology, JHU School of Medicine + Johns Hopkins Hospital; Department of Pathology, JHU SoM), Alani RM (co-corresponding; Department of Dermatology, Boston University). "The CoREST complex inhibitor, corin, decreases tumor growth, increases cellular differentiation and extends lifespan in ATRT xenograft models," *Neuro-Oncology* (advance access) 2026-07-21. DOI 10.1093/neuonc/noag160. PMID 42479134. **Science.** Atypical teratoid rhabdoid tumor (ATRT) — most common malignant brain tumor in infants — is driven by SMARCB1 loss, which cripples SWI/SNF and leaves the tumor dependent on the CoREST repressor complex. Corin is a bifunctional small molecule tethering an LSD1 inhibitor to an HDAC1/2 inhibitor in one molecule. It inhibited ATRT growth across every epigenetic subgroup, drove neuronal differentiation and apoptosis, increased chromatin accessibility at neuronal-differentiation/synaptic genes (ATAC-seq + RNA-seq), and suppressed orthotopic ATRT tumor growth with a significant extension of lifespan; on-target PD confirmed by increased H3K9ac and H3K4me1 in xenografts. Genetic validation: RCOR2 knockdown phenocopied corin and desensitized cells to the drug, confirming CoREST-complex specificity. **So what for Blackbird.** Enormous unmet need + real economics (a differentiation therapy for an infant brain tumor is exactly the profile that earns a rare-pediatric-disease priority review voucher, valuable on resale). BUT corin is not a JHU molecule — it is Philip Cole's compound, and Cole is now at Harvard/Brigham & Women's, so composition-of-matter sits elsewhere; what JHU owns here is the use — the CoREST-dependency-in-SMARCB1-null-tumors thesis and the ATRT method of treatment. That makes this a licensing-and-partnering conversation across two institutions more than a clean spin-out, and the broader LSD1-inhibitor field already has clinical-stage programs you would be racing. Intelligence note: co-author Jeffrey Rubens is now at Pfizer — signal on where pediatric-neuro-oncology talent/interest is flowing. Candid read: worth a look as a landscape + relationship item and a reason to keep a line open to the Raabe lab; the fractured IP means it is not the founding-a-company opportunity that Pick 1 is. Paper links: https://doi.org/10.1371/journal.pone.0353674 ; https://doi.org/10.1093/neuonc/noag160

Sourcing Radar — Fu lab JHU spinout Islex Therapeutics (Baltimore) islet-targeted ZnT8 antibody Isle43 reverses new-onset T1D in NOD mice in Diabetes 2026-07-20 (humanized Fc-silent beta-cell-homing antibody-as-chaperone: enhances ER folding, attenuates HLA-I, induces PD-L1 as PD biomarker; grant-stage Baltimore JHU spinout in the hot disease-modifying-T1D lane = time-sensitive Blackbird BioVentures co-investment sourcing lead) — dry preprint week otherwise, single-lead episode

Tue, 21 Jul 2026 12:00:00 +0000

Sourcing Radar for Tuesday, July 21, 2026. A dry JHU/UMB/Lieber preprint week (no corresponding-author commercializable preprint in the 2026-07-17→21 window), so a single strong Baltimore-company sourcing lead rather than off-thesis padding. **Pick 1 (LEAD, regional company + platform sourcing lead) — Fu lab JHU + Islex Therapeutics (Baltimore) islet-targeted ZnT8 antibody.** Guo Z (first), Kasinathan D, Yun S (Islex Therapeutics, LLC, Baltimore — CEO/co-founder), Golson ML (Division of Endocrinology, Diabetes & Metabolism, JHU SoM), Fu D (senior corresponding; Department of Physiology, Pharmacology & Therapeutics, Johns Hopkins School of Medicine). "Islet-Targeted ZnT8 Antibodies Protect Pancreatic β-Cells From Inflammatory Stress," *Diabetes*, published online 2026-07-20. DOI 10.2337/db26-0183. PMID 42474296. **Full-text note: ADA article is Cloudflare/subscription-gated with no preprint or PMC copy this run — new-paper results are from the published abstract; platform mechanism grounded in the fully-read 2023 foundational paper (PMC9876881).** **Platform (from 2023 paper, read in full).** Target is ZnT8 (zinc transporter 8, SLC30A8), a major T1D autoantigen presenting an extracellular epitope essentially only on the pancreatic beta-cell surface, cycled to the surface coupled to glucose-stimulated insulin secretion = a molecular zip-code for the beta cell. Fu lab cloned mAb43 against the native surface epitope: subnanomolar affinity (0.42 ± 0.05 nM), strictly conformation-specific, systemic dosing homes to pancreas (pancreas-to-serum ratio 24.6–66.2, no signal elsewhere); glucose stimulation raised ZnT8-mediated uptake ~31-fold; mAb43–IL-2-mutein fusion delivered immunomodulatory cargo islet-selectively. Core asset = a beta-cell-homing antibody chassis addressing the islet in vivo through the bloodstream without touching the rest of the body. **New paper (from abstract).** Humanized, Fc-silenced version "Isle43": internalized and acts as a selective molecular chaperone for ZnT8 inside the beta cell — enhances ER protein-folding capacity, attenuates HLA-I hyperexpression (less visible to autoimmune attack), and robustly induces PD-L1 (reinforces local immune checkpoint; used as a direct pharmacodynamic marker of on-target ZnT8-chaperoning). In vivo: durable pancreatic retention, dose-dependent reversal of new-onset T1D in NOD mice, sustained remission after treatment cessation, protection of human islet-graft function in vivo. 2024 predecessor already showed cell-surface ZnT8 antibody prevents and reverses autoimmune diabetes in NOD mice. Mechanism has matured across three papers from "mask and deliver" to "repair beta-cell proteostasis and flip the local immune switch." **So what for Blackbird.** Disease-modifying T1D is now real and hot: teplizumab (anti-CD3, Tzield) is approved and validated the premise but is a systemic immunosuppressant; the cell-replacement players (Vertex stem-cell-derived islets, Sana hypoimmune engineered islets) put up striking insulin-independence numbers but are transplant products requiring immunosuppression or complex engineering. Islex sits in an under-occupied lane — preserve/protect the patient's own beta cells using the islet's own surface address to deliver therapy locally, no systemic immunosuppression, no transplant. Humanized Fc-silent antibody + clean PD biomarker (PD-L1) = developable-looking profile; differentiated mechanism in a validated market. Islex Therapeutics: Baltimore JHU spinout, founded 2023, CEO/co-founder Shumei Yun; disclosed funding is NSF SBIR Phase I #2432114, $275,000 (2024-09-15→2025-08-31) = grant-stage with a humanized lead and essentially no institutional capital. Precisely the window Blackbird operates in — and the window a Baltimore program gets pulled to Boston/Bay Area if nobody local engages (cf. Aluco, Kynexis out-migration pattern this month). **Actions.** Eddie — time-sensitive Blackbird BioVentures sourcing conversation: is Islex raising, cap table + JHTV license structure, seed/co-investment fit, ideally with a BioHub bench to keep them in Baltimore. Diligence flag: 2023 platform paper declared no patents/no COI, so pin down exactly what ZnT8-antibody-chassis IP JHTV has filed and what is licensed to Islex vs held by the university. Avi/Yixuan — science diligence on the antibody-as-intracellular-chaperone novelty claim and the human-islet-graft data; does PD-L1 give a real translational readout into a first-in-human design. Add Dax Fu (JHU Physiology, Pharmacology & Therapeutics) and Shumei Yun (Islex) to the tracked list — founder-and-company pair worth a direct meeting. **Passed / watch:** Romero MR … Colas AR (senior corresponding, Sanford Burnham Prebys, La Jolla), "Sulfotransferase signaling sustains fibroblast identity and antagonizes therapeutic cardiac reprogramming," *Nat Commun* 2026-07-17, DOI 10.1038/s41467-026-75583-8, PMID 42469251 — druggable node (PIP4K2C inhibition enhances cardiac reprogramming / myocardial repair) with JHU Division of Cardiology co-authors (Kwon, Ranek, Murphy, Andersen) but corresponding author out-of-region and full text abstract-only this run; logged as watch, not featured. Paper links: https://doi.org/10.2337/db26-0183 ; https://pmc.ncbi.nlm.nih.gov/articles/PMC9876881/ ; https://www.islextherapeutics.com/ ; https://www.sbir.gov/awards/208358

Sourcing Radar — Brock lab JHU Sidney Kimmel + Bosmans (Ghent/VUB) NaV1.8 (SCN10A) channelopathy driving hereditary primary idiopathic hyperhidrosis in Sci Adv 2026-07-17 (genotype-stratified label-expansion lane for Vertex/Latigo/Merck NaV1.8 class + JHTV-owned AQP5 sweat-gland second-target lane) + Martin lab UMB Greenebaum MCAM/CD146 driving microtentacle homotypic clustering + reattachment in TNBC on TetherChip in Breast Cancer Res 2026-07-17 (activates existing UMB TetherChip + microtentacle patent portfolio around anti-metastatic drug-screening platform + FDA-approved vinorelbine repurposing readout)

Mon, 20 Jul 2026 12:00:00 +0000

Sourcing Radar for Monday, July 20, 2026. Two peer-reviewed papers that landed Thursday July 17 straight into top-tier journals. **Pick 1 (LEAD) — Brock lab JHU Sidney Kimmel + Bosmans Ghent/VUB NaV1.8/SCN10A hyperhidrosis channelopathy.** Yamauchi S (first), Ferrer-Pinós A, Wang J, Ahmed W, Aguti V, Chatterjee S, Weinerman I, Pierorazio PM, Kang GH, Sponseller PD, Chen Q, Elhai KA, Song Z, Guerrerio PA, Marsic D, Bosmans F (co-senior; Ghent University/VUB Brussel; frank.bosmans@ugent.be), Brock MV (senior corresponding; Departments of Surgery, Oncology, and Medicine, Sidney Kimmel Comprehensive Cancer Center, JHU School of Medicine; mbrock1@jhmi.edu). "A neurocutaneous NaV1.8 channelopathy underlies a genetic subtype of primary idiopathic hyperhidrosis," *Science Advances* 12(29):eaed3221, 2026-07-17. DOI 10.1126/sciadv.aed3221. PMID 42467787. PMCID PMC13378580. **COI: no competing interests declared.** Funding: NIH R01NS126398, 3UM1HG006542; FWO Vlaanderen; Horizon Europe MSCA-DN; AHA; Dysautonomia International; Banks Family Foundation. **Science.** WES on 32 multigenerational hyperhidrosis families — SCN10A variants enriched at 18.8% in cases vs 6.8% in controls (p=0.0155-0.0249). Lead variant p.R14L (arginine→leucine at pos 14) is a gain-of-function missense segregating with childhood-onset craniofacial/truncal hyperhidrosis. Biophysics: R14L doubles peak Na1.8 current density in ND7/23 (52→106 pA/pF, P<0.01) with -9 mV hyperpolarizing shift in activation V1/2; identical doubling in mouse Na1.8 R14L. NaV1.8 immuno-localized to human thoracic sympathetic ganglia + co-localized with ChAT+ postganglionic cholinergic sudomotor neurons. Sweat measurements in R14L knock-in: males +48.9±21.9% (P<0.001), females +50.3±18.7% (p=0.002). Reciprocal C1288W LoF variant produces anhidrosis phenotype (~35% reduction in KI mice). **Pharmacology:** NaV1.8-preferential inhibitor A-887826 at 10 mg/kg IP rescues sweating in R14L mice by 70.3±5.4% (males) / 68.6±4.8% (females), P<0.001. Reference-standard glycopyrrolate (72.8%), guanfacine (64.9%), CBD (53%). **Translational caveat authors flag: "clinically advanced human-optimized NaV1.8 inhibitor suzetrigine exhibits markedly reduced potency on mouse NaV1.8"** — mouse-A-887826 rescue does not automatically translate. AQP5 p.A193V surfaces as gland-level modifier in a subset of families (accelerated osmotic swelling + P<0.0001 permeability change at 72h) = second druggable target lane distal to sympathetic-neuron NaV1.8 axis. **So what for Blackbird.** Two commercial lanes. (a) **Genotype-stratified label-expansion for the NaV1.8 class**: Vertex suzetrigine (Journavx) approved in acute post-surgical pain + chronic-neuropathic-pain program running; Latigo Ph1 NaV1.8 blocker; Merck LTG-001. None have hyperhidrosis as declared indication + none have genotype-stratified enrichment strategy on the shelf. R14L carrier population is tractable Ph2 enrichment substrate (~200 patients, iodine-starch/gravimetry endpoint). Tuck-in-indication conversation for Vertex/Latigo/Merck. **BUT NaV1.8 composition-of-matter is Vertex + AbbVie owned; JHU/JHTV I-P is on the genotype-stratification method + biomarker patient-selection algorithm side.** (b) **AQP5 sweat-gland second-target lane** — gland-intrinsic druggable target NOT encumbered by Vertex/AbbVie composition-of-matter; JHTV-ownable target-directed I-P around AQP5 modulators in eccrine + potentially xerostomia/salivary + lacrimal-gland indications. Real Blackbird-fit sourcing conversation. **Actions.** Eddie — JHTV status conversation: does Brock/Bosmans have a companion filing on the genotype-stratified diagnostic + treatment-selection method (immediate license conversation with Vertex/Latigo/Merck), plus a stand-alone AQP5 target-directed I-P conversation. Hemaka — add Malcolm Brock (JHU Sidney Kimmel + Depts of Surgery, Oncology, Medicine) to tracked-PI list; he runs the JHU thoracic-sympathectomy program with direct access to the hereditary-hyperhidrosis patient cohort that would enroll into any genotype-stratified Ph2. **Pick 2 — Martin lab UMB Greenebaum MCAM/CD146 activates existing TetherChip patent portfolio around anti-metastatic drug-screening + vinorelbine repurposing in TNBC.** Annis DA (first), Thompson KN, Ju JA, Mull M, Gilchrist DE, Cornell JL, Omili DO, Vitolo MI (co-senior corresponding; UMB SoM Pharmacology & Physiology; mvitolo@som.umaryland.edu), Martin SS (senior corresponding; Marlene & Stewart Greenebaum NCI Comprehensive Cancer Center + Pharmacology & Physiology + Graduate Program in Molecular Medicine, UMB School of Medicine; ssmartin@som.umaryland.edu). "Melanoma cell adhesion molecule (MCAM) mediates microtentacle generation, homotypic clustering, and reattachment of non-adherent breast tumor cells," *Breast Cancer Research* 28, 2026-07-17. DOI 10.1186/s13058-026-02346-0. PMID 42469948. **COI verbatim: "The University of Maryland School of Medicine owns patents on the subject of microtentacles and microfluidic cell tethering that list one of the authors on the current manuscript (Stuart S. Martin) as an inventor."** Funding: METAvivor Foundation; NIH R01-CA154624, R01-CA124704 (Martin); NCI T32CA154274; NIGMS T32GM008181; Greenebaum P30-CA134274; MD Cigarette Restitution Fund CH-649-CRF. **Science.** MCF-10A OE (MCAM overexpression) + CRISPR MCAM-KO in TNBC lines. MCAM upregulates vimentin + acetylated α-tubulin (microtubule-stabilization PTM signature) + microtentacle (McTN) protrusions + homotypic cell clustering + reattachment on TetherChip substrate + migration on xCelligence RTCA. All phenotypes reverse in MCAM-KO. **Vinorelbine (FDA-approved vinca-alkaloid microtubule agent, on-market as generic for NSCLC) blocks all MCAM-driven microtentacle-supported phenotypes** = repurposing hypothesis testable with existing formulary compound. MCAM upstream of microtentacle-driven clustering-and-reattachment biology; mechanistic bridge is MCAM→microtubule cytoskeleton stabilization. **So what for Blackbird.** Not new IP — Martin lab has been building around microtentacle biology + TetherChip microfluidic platform for the last decade — but first paper to attach the patent portfolio to a defined molecular driver (MCAM), a defined disease (TNBC), and a repurposable FDA-approved drug that reverses the phenotype (vinorelbine). Three commercial threads: (a) **TetherChip microfluidic-cell-tethering platform** as licensable anti-metastatic drug-screening substrate — competitive landscape is Angle PLC Parsortix + Menarini Silicon Biosystems CellSearch (both enumeration only, do not resolve cluster/reattachment biology); TetherChip is substrate for anti-metastatic-drug screening, defensible category. UMB ownership on the record. (b) **MCAM/CD146 as anti-metastatic biologic target in TNBC** — existing antibody landscape (Prothena etc. in melanoma) but no current commercial program has the microtentacle-cluster-reattachment mechanism as anchor thesis in TNBC subset; target-directed conversation for UM Ventures track. (c) **Vinorelbine repurposing in MCAM-high TNBC** — window-opener investigator-initiated Ph2 at Greenebaum with TetherChip as PD biomarker; builds MCAM-anti-metastatic story with human data on formulary-drug economics. **Actions.** Yixuan/Avi — UM Ventures existing-patent-portfolio audit: current state of TetherChip + microtentacle patents, claims coverage (cell-tethering microfluidic device vs microtentacle drug-screening method), open licensing conversations with any diagnostics/biotech customer. Founding-conversation infrastructure with Stuart Martin — second decade at UMB, patent portfolio not yet anchored into Blackbird BioVentures-adjacent commercial vehicle. Third — MCAM-in-TNBC as target-directed I-P track (composition-of-treatment-method claim covering MCAM-driven anti-metastatic therapeutic development). **Framing.** Both picks landed same day (Thursday July 17), both peer-reviewed direct-to-final (no preprint runway), both in top-tier journals (Sci Adv + Breast Cancer Res). Brock/Bosmans is a genuine JHTV-legible sourcing lead on the method + AQP5-second-target sides. Martin MCAM/TetherChip is a UM Ventures existing-patent-portfolio activation conversation. Paper links: https://www.science.org/doi/10.1126/sciadv.aed3221 ; https://pmc.ncbi.nlm.nih.gov/articles/PMC13378580/ ; https://breast-cancer-research.biomedcentral.com/articles/10.1186/s13058-026-02346-0

Sourcing Radar — Fried lab JHU Chemistry circular-RNA looped-translation biomanufacturing platform for ultra-large fibrous proteins (bioRxiv 2026-07-17, active JHTV patent filing declared) + Deredge/Freel Meyers cross-UMB-JHU MInt-HDX hybrid physics/ML docking framework (bioRxiv 2026-07-16, COI clean)

Sun, 19 Jul 2026 12:00:00 +0000

Sourcing Radar for Sunday, July 19, 2026 (Sunday two-episode day; paired with Portfolio Watch). Two platform-IP sourcing leads that map directly onto the JHTV and UM Ventures × Blackbird BioVentures co-investment surface. **Pick 1 (LEAD) — Fried lab JHU Chemistry looped-translation ultra-large fibrous-protein biomanufacturing platform.** Xie Q (first), Papa LJ III, Barybin AM, Xiong M, Shoulders MD (MIT Chemistry), Fried SD (senior corresponding; sdfried@jhu.edu; Department of Chemistry, Johns Hopkins University). "Synthesis of Ultra-Large Fibrous Proteins from Bacteria via a Looped-Translation System," *bioRxiv* v1 2026.07.16.738376, posted 2026-07-17. **COI verbatim: "QX and SDF are filing for a patent related to techniques described in this work"** — active JHTV patent filing at time of preprint posting. Funding: NIGMS R35-GM161721 (Fried), NSF CAREER MCB-2045844, Sloan Foundation, Dreyfus Foundation, NIGMS R35-GM136354 (Shoulders MIT), NSF Graduate Fellowship. CC-BY-NC-ND 4.0. **Platform.** Circular-RNA-based ribosome translation system that turns linear mRNA into a covalently closed circular transcript — ribosome walks the loop iteratively, concatenating polypeptide until an exit signal terminates. Three engineering pieces: (1) synonymous-codon locker sequences suppress folding-and-stalling in repetitive tracts; (2) RNA-circularization chaperones promote loop formation across repetitive transcripts; (3) ribosome-traffic mathematical model identifies translation bottlenecks around the circular transcript for optimization. Generalizes across ≥6 classes of fibrous proteins (spidroin silk, tropoelastin, multiple collagen family members). Products exceed titin (largest known natural polypeptide) at 3.8 megadaltons. EM + bulk-material fabrication + mechanical analysis confirm ultra-high-molecular-weight material properties. Secretion coupled via programmed ribosomal frameshift as exit signal — polypeptide exported across membrane in both E. coli AND Bacillus subtilis (food-grade industrial-scale fermentation host). **So what for Blackbird.** Blackbird-audience-legible commercial addressable space is broader than the abstract's bioplastics/engineered-living-materials framing. Three biomedical applications sit inside immediate reach: (1) recombinant collagen matrices — wound-care, surgical mesh, tissue-engineering scaffolds currently filled with donor-tissue or animal-sourced collagen (Humacyte, Organogenesis, Integra Life Sciences); (2) recombinant elastin for vascular grafts, dermal fillers, elastomeric drug-delivery matrices (Elastagen → Allergan 2018 exit comp at >$100M valuation without ultra-high MW); (3) spider-silk sutures + surgical adhesives — defensible medical-materials space Bolt Threads + Spiber never entered because their platforms could not reach the required MW. Fried is a young senior faculty at JHU Chemistry with prior JHTV footprint; filing described as ongoing, JHTV had eyes on this pre-preprint. Single-institution JHU IP with MIT co-inventorship carve-out on Shoulders side. **Actions.** Eddie — JHTV patent-status conversation on field-of-use structure, composition-of-matter vs. method-claim separation, active licensing conversation with strategic partners (Modern Meadow, Ginkgo, industrial-fermentation majors). If no active licensing conversation, this is a founding-conversation moment with Fried; alongside Aletira as second fully-incubated BioHub NewCo model. Avi/Hemaka — biomanufacturing platform incubation conversation with Fried in near term. Diligence questions: process-scale fermenter titer (10L + 100L industrial yield), sequence-verified molecular-weight distribution + batch consistency for regulatory conversations, endotoxin/host-protein contamination profile (Bacillus subtilis preferred, and platform works in Bacillus), Fried founder disposition. **Pick 2 — Deredge/Freel Meyers cross-UMB-JHU MInt-HDX hybrid physics/ML docking framework.** Lowe V (first), Smith AK, Parakra R, Toci E, Freel Meyers CL (co-senior; JHU School of Medicine, Department of Pharmacology & Molecular Sciences), Deredge D (senior corresponding; UMB School of Pharmacy, Department of Pharmaceutical Sciences; dderedge@rx.umaryland.edu). "MInt-HDX: Leveraging Hydrogen-Deuterium Exchange Mass Spectrometry and Machine-Learning to Improve Protein-Ligand Docking," *bioRxiv* v1 2026.07.14.738285, posted 2026-07-16. **COI verbatim: "The authors have declared no competing interest."** Funding: NIH T32 GM158458 + R01GM143810 (Deredge). Cross-institution JHU + UMB — same UM Ventures × Blackbird co-investment surface as aSKY. **Platform.** XGBoost gradient-boosted-decision-tree machine-learning model trained on differential HDX-MS signatures across 11 protein-ligand systems (1,032 individual peptides). Predicts, for a new protein-ligand system with its own HDX-MS readout, which residues actually contact the ligand. Two-step docking workflow: (1) 3D clustering + convex-hull geometric step generates HDX-guided candidate docking sites (model tells physics-based docking engine where to look); (2) after physics-based docking generates candidate poses, XGBoost model filters + re-scores based on HDX-MS consistency. Validated on 3 systems, Ligand-RMSD <3 Å from crystallographic ligand conformation. Requires only solution-phase HDX-MS data, no co-crystal structure needed. **So what for Blackbird.** Licensable drug-discovery platform tool. Broad customer base — every major pharma runs an HDX-MS core (J&J, Pfizer, Merck, Roche, Novartis, AstraZeneca) and every serious structure-based-discovery biotech (Structure Therapeutics, Relay Therapeutics, Schrödinger, Nimbus) runs or contracts for HDX-MS. Value proposition: collapses HDX-MS run → diligence-legible pose model from weeks of manual interpretation to automated processing. Software platform with installed-base-driven revenue model. COI declared as none = JHTV + UM Ventures joint provisional-filing window open. Composition-of-workflow claims still patentable even if training code is public. Cross-institution JHU + UMB co-corresponding structure is the exact framework UM Ventures × Blackbird co-investment agreement supports — aSKY-track precedent. Second cross-institution JHU + UMB computational-tools paper in less than a month, alongside Popel-Fertig-Deshpande spatial-QSP paper (PNAS 2026-07-14) = pattern signal on where the co-investment surface should be spending time. **Actions.** Yixuan/Avi — JHTV + UM Ventures joint provisional-filing status on MInt-HDX composition-of-workflow claims; open-methods-plus-closed-workflow-composition IP posture check; Deredge + Freel Meyers founder-disposition read (commercial vehicle vs. academic-methods-only). Esther — add Deredge (UMB SoP Pharmaceutical Sciences) to tracked-lab list; pull patent history (Justia); pull Freel Meyers JHTV footprint separately. **Framing.** Sunday two-episode day (Portfolio Watch follows). Both picks are platform-IP sourcing leads with active or open JHTV/UM Ventures surface. Fried lab is the highest-value item — active JHTV patent filing declared in manuscript, JHU-led with single-institution IP posture and clear MIT co-inventor carve-out, addressable biomedical materials space is real. MInt-HDX is the cross-institution UM-Ventures pattern paper. Paper links: https://www.biorxiv.org/content/10.64898/2026.07.16.738376v1 ; https://www.biorxiv.org/content/10.64898/2026.07.14.738285v1

Portfolio Watch — FDA revokes Elevidys AAVrh74 platform-technology designation + OTP director search opens July 21 = AAV regulatory tightening around Aletira; Sensorion pivots SENS-501 OTOF → SENS-601 GJB2 easing hearing-loss density; GSK closes $10.6B Nuvalent acquisition July 15 reprices aSKY exit optionality; Nxera + Kynexis quiet loosens Lieber GPR52 calendar; Skyhawk FALCON-HD expands US/Canada/UK as translation-activator reference; Massive Bio TrialRelay direct competition to 1104health; JHU LSRI first proposal deadline July 15

Sun, 19 Jul 2026 12:30:00 +0000

Portfolio Watch for the week ending Sunday, July 19, 2026. Sunday two-episode day. **Week's biggest story = FDA revocation of Sarepta Elevidys AAVrh74 platform-technology designation + formal opening of CBER OTP permanent-director search (applications due Monday July 21)** — direct AAV regulatory-environment tightening around Aletira. First-ever revocation of an AAV "platform" designation, driven by Elevidys acute-liver-failure fatalities. Read for Aletira: platform-technology carve-out is not a durable review tool; each IND reviewed on own merits with fresh hepatotox + immunogenicity questions. SELEXON cell-type-restriction chassis (reduces off-target hepatocyte exposure) just became more valuable as a review-culture asset than as a chemistry-differentiation asset. Karim Mikhail acting CBER director covering OTP through at least Q3; interim leadership + tightened platform-designation posture = slower, more conservative AAV review environment for Aletira IND-enabling package. Not existential, real timeline drag. Watch-item. **Sensorion pivots from SENS-501 (OTOF) to SENS-601 (GJB2/connexin-26)** — modest tailwind hidden inside competitor pivot. OTOF space less crowded (Sensorion has stepped out of two-horse race against Regeneron/Akouos Otarmeni). GJB2 (second-most-common inherited hearing-loss target) validated as next commercially attractive white-space. Aletira SELEXON hair-cell-restriction chassis has natural extension into supporting-cell target set (GJB2 = cochlear supporting cells) — worth Geoffrey Lynn team's diligence. **GSK closes Nuvalent acquisition Tuesday July 15 at $10.6B** — zidesamtinib Sept 18 PDUFA + neladalkib Nov 27 PDUFA. GSK re-establishing oncology precision-medicine footprint post-Blenrep. Two reads for aSKY: strategic-acquirer side reprices what an NSCLC-adjacent KRAS-frame antibody-engineering platform could look like at exit for an early-stage program; competitive-landscape side sets ceiling for kinase-inhibitor space that any KRAS-frame antibody program has to navigate around. aSKY differentiation on antibody-engineering-versus-small-molecule intact but small-molecule end more crowded. **Adventris/Zaidi mKRAS-VAX interception paper landed Cancer Discovery Thursday July 16** — peer-reviewed publication of the first-in-human interception thesis (18/20 immunogenicity, TCR persistence to 2y, zero-progressor over median 16.5mo follow-up descriptive; already covered in Saturday Radar). Adventris formalized cancer-interception thesis for KRAS neoantigen space = sharpens modality-differentiation question (antibody vs peptide vaccine vs LN-targeting conjugate) for aSKY Series-A pitch. **Elicio $15M registered direct offering July 2** — competitor weakness signal (amphiphile-conjugate mutant KRAS vaccine competitor to Adventris; micro-cap raise to launch ELI-002 7P Ph1 = public-market signal amphiphile-conjugate has harder time attracting capital than peptide-vaccine story). **Lieber GPR52 calendar loosens** — Nxera earned $10M from separate AbbVie GPCR milestone but no NXE'149 partnership update in window. Kynexis KYN-5356 Ph2 still Q4 2026, no new update. No emraclidine + no new Cobenfy news = muscarinic space BMS-Cobenfy-dominant, GPR52 differentiation intact. Reverses last week's tighter-calendar read; Third Rock conversation now less about compressing DC timing and more about assembling full differentiation package (best-in-class chemistry + sex-stratified PD biomarker + multicellular tissue-state readout from Martinowich atlas infrastructure). **Slusher GCPII IBD quiet week** — no new tulisokibart or obefazimod data-set headlines. Vera atacicept (TRUTAKNA) IgAN accelerated approval July 7 continues to validate autoimmune-approval-velocity but no new direct IBD-mechanism entrant. Positioning stays what it was — differentiation on pain axis, defensibility on bimodal mechanism. **Skyhawk FALCON-HD expanded to US/Canada/UK** — >175 patients, >10 countries, >20 sites for pivotal HD splice-modulator. Citable reference point for modality narrative "small molecules can durably and safely modulate a target RNA in patients." Avi/Hemaka keep in outside-pitch deck for Blackbird translation-activator platform. CAMP4 routine inducement grant July 17, no material update on CMP-002 first-in-human. **Massive Bio launches TrialRelay** = physician-facing enrollment orchestration product directly at 1104health Shared Care Network category. OpenAI infrastructure partnership + 8-yr trial-matching corpus behind it. Rose needs to convert FDA RTCT pilot visibility into first-mover integration slot — final pilot selection criteria drop late July, pilots start August, AZ TrAVeRse MCL + Amgen STREAM-SCLC anchors already on Paradigm Health infrastructure. **JHU LSRI initial proposal deadline hit July 15** — first material submission event for $80M/yr internal translational-funding pool, applications opened June 15, Weeraratna committee oversight. Esther + Hemaka to work Weeraratna committee side for visibility onto submitted-project longlist; LSRI-miss-but-investor-current pipeline = exactly Blackbird sourcing pull. Durable tailwind now producing signal. Sources: https://www.fda.gov/vaccines-blood-biologics/tissue-tissue-products/elevidys ; https://endpoints.news/fdas-acting-director-of-cell-and-gene-therapies-to-leave-agency/ ; https://hearinghealthmatters.org/hearing-news-watch/2026/sensorion-gene-therapy-sens-601/ ; https://www.gsk.com/en-gb/media/press-releases/gsk-completes-acquisition-of-nuvalent-inc/ ; https://www.hopkinsmedicine.org/news/newsroom/news-releases/2026/07/experimental-kras-vaccine-generates-immune-response-against-pancreatic-cancer-in-people-at-high-risk ; https://www.globenewswire.com/news-release/2026/07/02/3321078/0/en/Elicio-Therapeutics-Announces-Pricing-of-15-Million-Registered-Direct-Offering.html ; https://www.prnewswire.com/news-releases/skyhawk-therapeutics-announces-expansion-of-its-global-pivotal-falcon-hd-clinical-trial-for-sky-0515-in-huntingtons-disease-to-the-united-states-canada-and-the-united-kingdom-302825053.html ; https://hub.jhu.edu/2026/06/15/research-funding-opportunities-application-period/

Sourcing Radar — Zaidi mKRAS-VAX first-in-human pancreatic-cancer interception readout (Cancer Discovery 2026-07-16) + Corti UMB cytoglobin-sGC hypoplastic-heart druggable axis (bioRxiv 2026-07-16) + JHU-CS EnzGFM enzyme foundation-model landscape signal (Nat Commun 2026-07-17): Pick 1 (LEAD) Haldar SD first + Zaidi N senior corresponding, JHU Medicine + Sidney Kimmel Comprehensive Cancer Center + Bloomberg-Kimmel Institute for Cancer Immunotherapy, Cancer Discovery 2026-07-16 (DOI 10.1158/2159-8290.CD-25-2245, PMID 42458705, trial NCT05013216), "First-in-human testing of a mutant KRAS vaccine for pancreatic cancer interception in high-risk cohorts" — n=20 hereditary-predisposition individuals (BRCA1/BRCA2/PALB2/Lynch/FPC kindreds) with radiographic pancreatic abnormality on surveillance imaging received pooled six-peptide mKRAS-VAX; grade 1-2 AEs only, 18/20 (90%) mounted mutant-KRAS-specific T-cell response, TCR-sequencing shows vaccine-induced clonotypes persist up to 2y, zero progression to PDAC over median 16.5-mo follow-up (descriptive not powered); first solid-tumor cancer-interception vaccine paper — payer-driven prevention economics + germline-defined population; regulatory template for the AACR-Cancer Interception Task Force pathway; Adventris Pharmaceuticals (YC-backed JHU spinout; Jaffee co-founder + CSO; Zaidi + Yarchoan equity) is the commercial vehicle — pipeline expansion beyond mCRC/adjuvant PDAC into interception is a distinct clinical development plan + distinct Series-B financing conversation; direct competitive read for aSKY on KRAS-frame antibody-engineering vs peptide-vaccine vs Elicio LN-targeting-conjugate modality positioning. Pick 2 Clark A first + Corti P senior corresponding, UMB SoM Biochemistry & Molecular Biology + Center for Blood Oxygen Transport and Hemostasis, bioRxiv v1 2026-07-16 (DOI 10.64898/2026.03.13.711730), "Cytoglobin-dependent nitric oxide/soluble guanylate cyclase signaling is required for ventricular development and function" — zebrafish cygb-KO disrupts NO-sGC-cGMP signaling exactly during cardiac progenitor migration in anterior lateral plate mesoderm, producing small compact thick-walled ventricle with reduced diastolic filling + reduced stroke volume that recapitulates HLHS; sGCα1 KO phenocopies cygb mutant; two pharmacological rescues both work — NO donor and Bayer's Cinaciguat sGC activator (previously Ph2 acute decompensated heart failure) fully restore stroke volume and normalize ventricular wall thickness when given during the critical early developmental window; explains why inhaled NO stabilizes HLHS neonates clinically + opens congenital-heart-disease indication for the sGC activator/stimulator class (Merck Verquvo vericiguat + Bayer Adempas riociguat + industry-wide next-gen sGC medchem); COI verbatim: none; funding NIH T32AR7592 + NHLBI R01 NHL168775; mechanism-first paper, no patent/company disclosed — IP-white-space window for UM Ventures provisional-filing conversation; second UMB Shock-Trauma-and-Anesthesiology-adjacent cardiovascular druggable-axis paper this week after Chao TLR7-ITGAM ischemia-reperfusion — UMB cardiovascular translational bench is accumulating fast, worth flagging in monthly UM Ventures partnership sync. Pick 3 (landscape signal + diligence framing, not a sourcing lead) Wang C first (Henan Medical University), Yu Y + Wang T co-corresponding (Henan Medical), Shen Y co-corresponding (JHU Computer Science, yshen92@jhu.edu), Nat Commun 2026-07-17 (DOI 10.1038/s41467-026-75283-3, PMID 42469205), "EnzGFM: a genomic foundation model for enzymes with hierarchical pre-training and agent-based reasoning" — Mamba-Transformer hybrid enzyme-specific PLM with hierarchical pre-training beats standard PLMs by 13-20% across four benchmarks (kinetic parameters, enzyme-reaction mapping, EC-number classification, mutation-effect assessment) at 2-5x inference throughput; ships with EnzGFM-Agent agentic wrapper for beneficial-variant enrichment; 8/9 authors at Chinese institutions (Henan Medical + Shanghai Jiao Tong + Shanghai AI Laboratory), only Shen at JHU-CS, funding entirely Chinese (NSFC + Henan Provincial + HAST), no patents disclosed; not a JHU-anchored sourcing lead, but the benchmark set becomes a public reference point every AI-first protein-design pitch (Cyrus + Xaira + Chai Discovery + EvolvePro + Basecamp + Ginkgo) should be pressed on for enzyme-design differentiation claims — SOTA in enzyme-foundation-models now landing in Nat Commun from a mostly-Chinese consortium is a competitive-dynamics scouting datapoint, directly touches Aletira SELEXON extension conversations.

Sat, 18 Jul 2026 12:00:00 +0000

Sourcing Radar for Saturday, July 18, 2026. Three items: a Johns Hopkins first-in-human milestone in cancer-interception vaccinology, a UMB mechanism paper opening a new indication for a clinical-stage drug class, and an enzyme foundation-model paper best read as diligence framing rather than as a sourcing lead. **Pick 1 (LEAD, direct JHU immuno-oncology franchise read) — Neeha Zaidi + Liz Jaffee report the phase-one, first-in-human readout for the pooled six-peptide mutant-KRAS vaccine used not for treatment but for cancer interception.** Haldar SD (first; MD Anderson, formerly Johns Hopkins), Huff AL, Wang HH, Zhu Z, Berg M, Lu J, Sun N, Abou Diwan E, Sinan H, Thoburn CJ, Guo MZ, Yoshida T, Chu LC, Ferguson AK, Sidiropoulos DN, Kagohara LT, Ho WJ, Bever KM, Baretti M, Yarchoan M, Laheru DA, Nauroth JM, Thomas AM, Wang H, Azad NS, Goggins MG, Jaffee EM, Zaidi N (senior corresponding; Johns Hopkins Medicine + Sidney Kimmel Comprehensive Cancer Center + Bloomberg-Kimmel Institute for Cancer Immunotherapy). "First-in-human testing of a mutant KRAS vaccine for pancreatic cancer interception in high-risk cohorts," *Cancer Discovery* 2026-07-16, DOI 10.1158/2159-8290.CD-25-2245, PMID 42458705. Trial NCT05013216. **Population.** Twenty individuals carrying hereditary pancreatic-cancer predisposition variants — germline BRCA1, BRCA2, PALB2, Lynch syndrome, familial pancreatic-cancer kindreds — plus a radiographic pancreatic abnormality on surveillance imaging. Not treatment population — these are people at high risk who have not yet developed PDAC, where current surveillance detects only a minority of precursor lesions early enough to intervene. **Result.** Primary safety + immunogenicity endpoints both cleared. AEs grade 1-2 only. 18/20 (90%) mounted mutant-KRAS-specific T-cell response. Longitudinal TCR-sequencing shows vaccine-induced clonotypes persist up to 2 years after priming — the durability signal the peptide-plus-adjuvant field has been waiting for. Zero participants progressed to PDAC over median 16.5-month follow-up (descriptive not powered — n=20, but the direction is right and the mechanism coherent). **So-what for Blackbird.** The paradigm shift + the spinout. Paradigm shift: first solid-tumor cancer-interception vaccine paper. Interception is a different market — payer-driven prevention economics not oncology-clinic infusion revenue, addressable population defined by germline sequencing not tumor staging, ten-to-fifteen-year natural history. Durable T-cell persistence from a peptide vaccine in a well-defined high-risk cohort gives the AACR-Cancer Interception Task Force pathway its first-mover regulatory template. Spinout: **Adventris Pharmaceuticals** — YC-backed JHU commercialization vehicle for the entire mKRAS-VAX platform (Jaffee co-founder + CSO; Zaidi + Yarchoan equity). Already framed on treatment programs across metastatic CRC and adjuvant pancreatic; this paper is the pipeline expansion into interception, a distinct clinical-development plan + distinct commercial thesis + distinct Series-B investor set. Not a Blackbird company, but the JHU-licensed asset that most directly touches neoantigen-vaccine and cancer-interception conversations. Action for Eddie: ping Adventris + JHTV oncology desk on whether the interception franchise is being carved into a separate development plan and financing conversation. Read-through for aSKY: Krupnick's KRAS-frame antibody-engineering platform now sits in the same target space as a phase-one-validated peptide-vaccine competitor and a growing set of amphiphile-conjugate vaccines out of Elicio — the modality-differentiation question (antibody-engineering vs peptide vaccine vs LN-targeting conjugate) is sharpened by every new mKRAS-VAX + Elicio data point, needs to be legible in aSKY's Series-A pitch. **Pick 2 (UM Ventures × Blackbird sourcing surface, mechanism-first white-space) — Paola Corti UMB Biochemistry opens congenital heart disease for the sGC activator/stimulator class.** Clark A (first), Corti P (senior corresponding; UMB SoM Department of Biochemistry and Molecular Biology + UMB Center for Blood Oxygen Transport and Hemostasis), Iqbal M, Hejlesen R, Weng T-T, Bruce A. "Cytoglobin-dependent nitric oxide/soluble guanylate cyclase signaling is required for ventricular development and function," bioRxiv v1 2026.03.13.711730, posted 2026-07-16, DOI 10.64898/2026.03.13.711730. **COI: none.** Funding: NIH T32AR7592 (training grant) + NHLBI R01 NHL168775 (Corti). Institutional oversight: UMB IACUC-1022015. **Setup.** Hypoplastic left heart syndrome — lethal congenital heart disease, small compact underdeveloped left ventricle in newborns, few thousand US babies per year, no drug therapy for the underlying malformation. Neonatologists give inhaled nitric oxide to stabilize cardiopulmonary function; the field has known iNO helps but not whether NO signaling is contributing to underlying pathology or providing symptomatic support only. **Mechanism.** Cytoglobin — heme protein previously thought to just scavenge NO — actually enhances NO signaling through the canonical soluble-guanylate-cyclase-to-cyclic-GMP pathway during embryonic heart development. Zebrafish cygb knockout disrupts NO signaling exactly during cardiac progenitor migration in the anterior lateral plate mesoderm; cardiac progenitors fail to converge and compact properly; ventricle becomes smaller, more compact, thicker-walled; reduced diastolic filling + reduced stroke volume — recapitulates the human hypoplastic phenotype. Genetic knockout of soluble guanylate cyclase α1 subunit phenocopies the cygb mutant — confirms the downstream effector axis. **Pharmacology beat.** Two rescues both work. (1) NO donor rescues stroke volume and ventricular architecture in cygb mutants. (2) **Bayer's Cinaciguat — clinical-stage sGC activator, previously Ph2 in acute decompensated heart failure — fully restores stroke volume and normalizes ventricular wall thickness when given during the critical early developmental window.** **So-what for Blackbird.** The sGC activator/stimulator class is now a real clinical franchise (Merck Verquvo/vericiguat approved for HFrEF; Bayer Adempas/riociguat approved for PAH; active industry medchem effort on next-gen sGC molecules). This paper opens a completely new indication — hypoplastic ventricular disease and congenital heart disease broadly — for the class, with a mechanism narrative that connects bedside iNO in HLHS neonates to a molecular pathway the class already engages. Mechanism-first paper, no patent/company disclosed — IP-timing pattern is exactly the moment a translational venture office should be having a provisional-filing conversation before somebody else does. Two questions for the team. **Avi + Hemaka:** worth having Esther or Yixuan reach out to Corti's group to ask what the current UM Ventures IP posture is on the cytoglobin-sGC-in-HLHS axis, and whether Corti has already had exploratory conversations with Bayer or Merck on repositioning their sGC estates into congenital heart disease. **Second beat:** this is our second UMB Shock-Trauma-and-Anesthesiology-adjacent cardiovascular druggable-axis paper this week, coming off the Chao lab TLR7-ITGAM ischemic-reperfusion paper Wednesday. UM Ventures has a mini-portfolio of translationally-mature cardiovascular biology accumulating quickly — pattern worth flagging in the next monthly with the UM Ventures partnership team. **Pick 3 (landscape signal + diligence framing, tail read) — EnzGFM, an enzyme-specific protein language model out of a mostly-Chinese consortium with a Johns Hopkins Computer Science co-corresponding author.** Wang C (first; Henan Medical University), Li M, Geng S, Li W, Zhou X, Wang YG, Yu Y (co-corresponding; Henan Medical), Wang T (co-corresponding; Henan Medical), Shen Y (co-corresponding; Johns Hopkins University Department of Computer Science, yshen92@jhu.edu). "EnzGFM: a genomic foundation model for enzymes with hierarchical pre-training and agent-based reasoning," *Nature Communications* 2026-07-17, DOI 10.1038/s41467-026-75283-3, PMID 42469205. **COI: none.** Funding: NSFC + Henan Provincial programs + HAST sponsorship. No patents disclosed. **Model.** Mamba-Transformer hybrid architecture, enzyme-specific pre-training corpus, hierarchical pre-training strategy. Beats standard protein-language-model baselines by 13-20% across four benchmarks — kinetic-parameter prediction, enzyme-reaction mapping, EC-number classification, mutation-effect assessment — at 2-5× the inference throughput of transformer-only baselines. Ships with EnzGFM-Agent, an agent-based reasoning wrapper that enriches beneficial enzyme variants in small experimental candidate pools. **Authorship + IP posture.** Nine authors, eight at Chinese institutions (Henan Medical + Shanghai Jiao Tong + Shanghai AI Laboratory), one at JHU. All three co-corresponding-author slots on the Chinese side. Funding entirely Chinese. No patents disclosed. Composition-of-workflow IP posture will be led from the Chinese authorship side — this is not the JHU-anchored sourcing conversation the Barrick protein-design paper in *Nature Chemical Biology* last week was. **So-what for Blackbird — diligence framing not a sourcing lead.** Every AI-first protein-design pitch cohort walking through Blackbird (Cyrus, Xaira, Chai Discovery, EvolvePro, increasingly Basecamp Research and Ginkgo) is claiming platform advantages in enzyme design specifically. The EnzGFM benchmark set now becomes a public reference point any pitch team should be pressed on: if the platform cannot beat EnzGFM on the same four public benchmarks or explain why its enzyme-specific pre-training data is meaningfully differentiated, that is a diligence red flag. Second beat — SOTA in enzyme-specific foundation models landing in Nat Commun from a mostly-Chinese consortium is itself a competitive-dynamics scouting datapoint on where enzyme-engineering foundation-model development is heading; the same trajectory is coming for antibody + effector-protein design and directly touches Aletira's SELEXON extension conversations. Paper links, trial number, and Adventris Pharmaceuticals corporate detail in the show notes.

Sourcing Radar — JHU-anchored three-picker with the strongest direct sourcing lead of the month: Pick 1 (LEAD) Pan D first + Xu Cao senior corresponding, JHU Orthopedic Surgery + BME/Institute for Cell Engineering, Nat Commun 2026-07-14 (DOI 10.1038/s41467-026-75423-9, PMID 42448693), "Dorsal horn DCC amplification loop induced by endplate osteoclasts generates chronic nociplastic low back pain in male mice" — endplate osteoclasts secrete netrin-1 that activates DCC on Advillin+ DRG neurons; DCC induces Src phosphorylation + SNARE docking + presynaptic glutamate release in the dorsal horn; DH neurons then upregulate netrin-1/DCC themselves as postsynaptic plasticity, closing a self-amplifying feedback loop that extends up to parabrachial nucleus; conditional netrin-1 KO in Trap+ osteoclasts + Dcc KO in Avil+ DRG both reduce hypersensitivity; effect replicates in naturally-aged mice; first mechanistic axis for non-specific chronic LBP (world's leading cause of years lived with disability, no MOA-defined therapy); two clean intervention points — netrin-1/DCC-directed biologic + osteoclast-endplate-targeted small molecule; COI verbatim "the authors declare no competing interests"; single-institution JHU IP; NIH NIA-funded (P01AG066603 + R01AG076783 + R01AG068997 all to X.C.) — direct JHU sourcing lead + JHTV provisional-conversation moment; sex-specificity caveat (male mice) is next-cycle diligence question. Pick 2 (platform) Su Y first + Hai-Quan Mao senior corresponding, JHU Institute for NanoBioTechnology + BME + Materials Science + Translational Therapeutic and Regenerative Engineering Center, ACS Appl Mater Interfaces 2026-07-16 (DOI 10.1021/acsami.6c06888, PMID 42461776), "Stabilizing Anionic mRNA Lipid Nanoparticles by Cleavable Crosslinking of Cholesterol" — bolt-on acid-cleavable cholesterol crosslinker + PEG-diamine stabilizes anionic LNPs without swapping primary lipid constituents, closes the anionic-corner-of-design-space product problem behind splenic-biased mRNA delivery; optimized crosslinked formulation delivers significant increase in splenic mRNA expression at 12h + 33.2% of CD45+tdTomato+ splenic cells identified as T cells (transduction of lymphocytes); Hopkins-only manuscript; enables tolerogenic mRNA for autoimmunity + in-vivo mRNA-CAR-T (skips ex-vivo T-cell engineering step commercial CAR-T programs are struggling to make economically viable) + T-cell-directed cancer vaccines; Mao lab has serial NIBIB commercial spinout history — platform IP with real optionality across multi-billion-dollar therapeutic areas; natural fit with the Luetkens UMB cathepsin-B-CAR-T story (in-vivo mRNA-CAR-T formulation chassis pairs with adjunct that preserves CAR-T persistence). Pick 3 (landscape + intelligence surfacing) Radulescu E first + co-corresponding + Weinberger DR senior corresponding, Lieber Institute for Brain Development + JHU Psychiatry + Genetic Medicine + Neurology + Neuroscience, Nat Commun 2026-07-16 (DOI 10.1038/s41467-026-75470-2, PMID 42463512), "Analysis of gene co-expression connectivity dynamics implicates aberrant neuron-oligodendroglia interactions in schizophrenia" — bulk RNA-seq DLPFC + hippocampus + caudate (n=297/250/349), differentially connected genes converge on neuron-oligodendroglia interaction pathways confirmed in snRNA-seq + iPSC brain organoids; directional validation of Martinowich 2026-07-14 multicellular-schizophrenia framing for the Lieber neuropsych NewCo GPR52 DC-declaration biomarker strategy; COI surfaces Aluco BioSciences (San Carlos CA; CA-incorporated 2024-03-24) as new Lieber-adjacent West Coast neurology + endocrine biotech via Jennifer Erwin employee-with-equity disclosure — Erwin was Lieber-based staff scientist during study (funded MSCRF Launch Program 2023-MSCRFL-6185); Erwin's current affiliation only Aluco (not Lieber) — Aluco added as new tracked-company entry for Blackbird BioVentures + Chief Business Officer first-read on IP/programs. Second beat — pattern of Baltimore-discovery neuropsych science leaving Baltimore for Series-A commercial vehicles (Aluco to San Carlos + Kynexis to Netherlands from UMB KAT-II) is the leakage the UM Ventures × Blackbird co-investment surface is designed to prevent — Lieber-side equivalent needed structurally. Bullmore also discloses Nxera Pharma (Sosei Heptares; direct GPR52 competitor to Lieber NewCo) consulting relationship.

Fri, 17 Jul 2026 12:00:00 +0000

Sourcing Radar for Friday, July 17, 2026. Johns Hopkins-anchored three-picker: strongest direct sourcing lead of the month + a splenic-T-cell-targeting LNP platform paper + a Lieber Nature Communications discovery paper that surfaces a new California neurology biotech (Aluco BioSciences) via COI disclosure. **Pick 1 (LEAD, direct sourcing lead) — Xu Cao lab at Johns Hopkins Orthopedic Surgery assembles the first mechanistically-anchored, druggable axis in non-specific chronic low back pain.** Pan D (first), Shen M, Abatan E, Zheng J, Nasser P (Mount Sinai Orthopedics), Laudier DM, Kong C, Yan L, Qiu Z, Iatridis JC, Wan M, Zheng J, Taylor BK (Pittsburgh Center for Pain Research), Cao X (senior corresponding; xcao11@jhmi.edu). "Dorsal horn DCC amplification loop induced by endplate osteoclasts generates chronic nociplastic low back pain in male mice," *Nature Communications* 2026-07-14, DOI 10.1038/s41467-026-75423-9, PMID 42448693. Corresponding author at Department of Orthopedic Surgery + Department of Biomedical Engineering and Institute of Cell Engineering, Johns Hopkins University School of Medicine. **Mechanism.** Spine degeneration → porous endplates → Trap+ osteoclasts inside porous endplates secrete netrin-1 → activates DCC on Advillin+ DRG sensory neurons → DCC triggers Src phosphorylation + SNARE-complex docking → presynaptic glutamate release into the dorsal horn → DH neurons upregulate netrin-1/DCC themselves as postsynaptic plasticity → self-amplifying positive feedback loop → extends to parabrachial nucleus (ascending pain-affect circuit). Conditional netrin-1 KO in Trap+ osteoclasts + Dcc KO in Avil+ DRG neurons both reduce spinal hypersensitivity — both genetic legs are causal. Effect replicates in naturally-aged mice (not just lesion-model artifact). **IP + COI.** Verbatim: "the authors declare no competing interests." Funding exclusively NIH NIA (P01AG066603, R01AG076783, R01AG068997 all to X.C.). Single-institution JHU IP (with Mount Sinai + Pitt co-authors). This is the "no licenses yet, NIH grants paying for it, single-institution" fact pattern Blackbird wants when a JHTV provisional-filing conversation is the natural next step. **So-what for Blackbird.** Strongest JHU direct sourcing lead the Sourcing Radar has surfaced this month. Non-specific chronic LBP is the world's leading cause of years lived with disability, no mechanism-defined therapy, one-in-six-adults patient population. Two clean therapeutic intervention points fall out: (1) netrin-1 or DCC-directed biologic (NetrisPharma of Lyon has anti-netrin-1 NP137 in oncology trials — antibody path against netrin-1 is precedented, but not in pain indication); (2) osteoclast-endplate-targeted small molecule — sits inside the bone-therapeutics chemistry space where Cao lab and JHU Ortho has 15-year IP track record. Action for Eddie + Hemaka this week: JHTV conversation on provisional-filing status + field-of-use structure + Cao founder disposition. Diligence questions cluster in three places: (a) sex-specificity — paper is male-mice-only, next-cycle question; (b) human tissue validation of netrin-1/DCC axis in patient endplates + DRG; (c) competitive intelligence around the D-C-C or osteoclast-endplate-targeted white space. **Pick 2 (platform) — Hai-Quan Mao lab bolt-on cleavable-crosslinked cholesterol chemistry stabilizes anionic LNPs for splenic-biased T-cell mRNA delivery.** Su Y (first), Choy J, Liu X, Zhu Y, Lin J, Wei C, Goodier KD, Yu D, Cheng L, Patel M, Lu X, Ma J, Wang J, Mao HQ (senior corresponding; hmao@jhu.edu). "Stabilizing Anionic mRNA Lipid Nanoparticles by Cleavable Crosslinking of Cholesterol," *ACS Applied Materials & Interfaces* 2026-07-16, DOI 10.1021/acsami.6c06888, PMID 42461776. Hopkins-only manuscript — Institute for NanoBioTechnology + Department of Biomedical Engineering + Department of Materials Science and Engineering + Translational Therapeutic and Regenerative Engineering Center. **Platform.** Acid-cleavable cholesterol crosslinker + PEG-diamine adds a stabilization layer to any anionic LNP formulation without swapping the primary lipid constituents — closes the actual product problem behind why anionic LNPs are not commercial (anionic helpers redirect biodistribution to spleen but weaken RNA-lipid interactions during purification, so nanoparticles fall apart; every anionic-LNP program hits this ceiling). Acid-cleavable means tether releases in the endosome so payload release is preserved. Crosslinker length is tunable — chemistry space for follow-on medicinal-chemistry programs. **Results.** Optimized crosslinked LNP produces significant increase in splenic mRNA expression at 12h vs uncrosslinked control. 33.2% of splenic CD45+ tdTomato+ cells identified by flow are T cells — payload is landing in the spleen AND transducing lymphocytes. Mechanistic analyses point to controlled mRNA release + altered intracellular processing. **So-what for Blackbird.** Platform IP wraps around every downstream therapeutic program that wants to touch the anionic corner of LNP design space. Splenic-biased T-cell mRNA is the workhorse formulation target for (a) tolerogenic mRNA for autoimmunity (Cabaletta, Vor Bio, Kernel — active competitive space); (b) in-vivo mRNA-CAR-T (Umoja, Capstan, Orbital Therapeutics — the market Cartesian and Moderna are racing toward as commercial CAR-T economics wobble); (c) T-cell-directed cancer vaccines at splenic marginal zone. Natural sourcing conversation: JHTV status on the crosslinker composition + tunable-length claim + Mao founder disposition. Direct portfolio adjacency to Luetkens UMB cathepsin-B-CAR-T story — in-vivo mRNA-CAR-T formulation chassis pairs with adjunct that preserves CAR-T persistence. **Pick 3 (landscape + intelligence surfacing) — Radulescu-Weinberger Lieber Institute Nature Communications discovery paper on schizophrenia is directionally consistent with Martinowich multicellular framing, but the Blackbird-relevant surface is the COI disclosure of Aluco BioSciences as a Lieber-adjacent West Coast neurology + endocrine biotech.** Radulescu E (first + co-corresponding; eugenia.radulescu@libd.org), Vértes PE (Cambridge Psychiatry), Han S, Erwin JA (Aluco BioSciences, San Carlos CA), Paquola ACM (JHU Neurology → DataTecnica/NIA CARD), Hyde TM, Kleinman JE, Sawada T, Bullmore ET, Weinberger DR (senior corresponding; drweinberger@libd.org). "Analysis of gene co-expression connectivity dynamics implicates aberrant neuron-oligodendroglia interactions in schizophrenia," *Nature Communications* 2026-07-16, DOI 10.1038/s41467-026-75470-2, PMID 42463512. **Science.** Bulk RNA-seq DLPFC (n=297) + hippocampus (n=250) + caudate (n=349); differentially connected genes across three network metrics; enrichment in neuron-oligodendroglia interaction pathways; confirmed in snRNA-seq + four independent iPSC brain organoids. Myelination-connectivity story of schizophrenia. Directional validation of Martinowich CSO 2026-07-14 multicellular-schizophrenia framing (§2 Neuropsych NewCo entry) for the Lieber NewCo GPR52 DC-declaration biomarker strategy. **COI intelligence (verbatim).** "Daniel R. Weinberger serves on the Scientific Advisory Boards Sage Therapeutics and Pasithea Therapeutics. Edward T. Bullmore has consulted for Boehringer Ingelheim, SR One, GlaxoSmithKline, Sosei Heptares and Monument Therapeutics. Petra E. Vértes consults for LinusBio Ltd. **Jennifer A. Erwin is an employee of Aluco Biosciences and holds equity in the company.** J.E.K. is a member of a drug monitoring committee for an antipsychotic drug trial for Merck. All other authors declare no competing interests." **Aluco BioSciences facts.** California-incorporated 2024-03-24, 733 Industrial Rd, San Carlos CA 94070 (Bizapedia + California SoS). Website (alucobio.com) is a soft-launch placeholder — describes focus as neurological and endocrine disease areas, "novel approach to drug development" and "compounds tailored to specific disease indications." No public team, target, or funding disclosure yet. Erwin's only affiliation on this paper is Aluco (not Lieber), but funding statement confirms Erwin was Lieber-based during the study (Maryland Stem Cell Research Fund Launch Program 2023-MSCRFL-6185, a Lieber/JHU-anchored mechanism). **So-what for Blackbird.** Intelligence, not a sourcing lead. Two implications: (1) Add Aluco BioSciences to tracked-company list; Chief Business Officer first-read on what Aluco is actually developing — publicly the website is a placeholder, so the intel channel is direct. Three questions to answer: is Aluco licensing Lieber/JHU IP; does Weinberger have any relationship beyond Sage + Pasithea; is this a schizophrenia-adjacent asset in same design space as Lieber GPR52 NewCo. (2) Pattern signal — two Lieber/UMB-adjacent neuropsych assets have now left Baltimore for Series-A commercial vehicles (Aluco to San Carlos, Kynexis to Netherlands from UMB KAT-II biology). That is the leakage the UM Ventures × Blackbird co-investment surface is designed to prevent — Lieber-side equivalent structurally needed. Worth Matt raising with Lieber board. Second data-point: Bullmore also discloses Nxera Pharma (Sosei Heptares) consulting — Nxera is the direct GPR52 same-mechanism competitor to Lieber NewCo we flagged on 2026-07-12. Full paper links and Aluco BioSciences address in the show notes.

Sourcing Radar — UMB-anchored three-picker with a filed methods-of-use PCT as the lead: Pick 1 Li Y first + Wei Chao senior corresponding UMB SoM Anesthesiology + Center for Shock, Trauma and Anesthesiology Research JACC Basic Transl Sci 2026-07-14 (DOI 10.1016/j.jacbts.2026.101626, PMID 42447539) "TLR7 Inhibition Limits Ischemic Cardiac Injury by Disrupting ITGAM-Dependent Immune-Endothelial Interactions" — mouse LAD-ligation IR model (45 min ischemia + 24 h reperfusion) shows TLR7 KO + pharmacological inhibition with enpatoran (Merck KGaA TLR7/8 antagonist in SLE/CLE/myositis autoimmune trials) cut infarct-over-area-at-risk by 35-76% across prevention/treatment/early-rescue dosing windows with epistasis to genetic KO confirming target; single-nucleus RNA-seq localizes mechanism to TLR7-driven inflammatory endothelial + coordinated myeloid program converging on ITGAM (CD11b) integrin upregulation; anti-ITGAM antibody (Bio X Cell BE0007) blockade phenocopies TLR7 KO — MI reduction + reduced neutrophil/monocyte infiltration; PCT/US2025/037133 "Methods of Treating Ischemia/Reperfusion Injury" filed as UMB IP — enpatoran composition-of-matter is Merck KGaA's but methods-of-use for cardio-reperfusion is UMB's, cross-licensing leverage; funded R35-GM140822 (Chao) + AHA 26POST1559348 (Li); direct UM Ventures × Blackbird BioVentures co-investment sourcing lead. Pick 2 platform paper Wang J + Liang Y senior corresponding UMB SoM Diagnostic Radiology (with Walczak P co-supervisor at UMB + K. Wang JHU BME + J. Xu Kennedy Krieger F.M. Kirby Center for Functional Brain Imaging) Adv Sci 2026-07-14 (DOI 10.1002/advs.202515913, PMID 42447126, PMC13367743 open access) "Intravital Multimodal Imaging of Human Cortical Organoid Transplantation in a Mouse Model of Chronic Stroke" — MIPOT (Multimodal Imaging Platform for Organoid Tracking) integrates 9.4T MRI + Luc2-mCherry bioluminescence + intraoperative surgical microscopy + intravital two-photon multicolor calcium imaging on one animal via 3D-printed removable head-fixation adapter; hiPSC (KOLF2.1J) cortical organoids grafted into cleared photothrombotic-stroke cavities in male C57BL/6J mice; bioluminescence drops to ~50% by day 4, ~25% by day 7, ~5% by day 14 with limited human-graft persistence at endpoint histology — honest platform paper, no efficacy claim, no COI, no patent, licensable cross-institution UMB+JHU+KKI imaging chassis for CNS cell-therapy diligence. Pick 3 (framing, tail item) Todd Gould first UMB SoM Psychiatry + Pharmacology-Physiology and Drug Discovery + Neurobiology (co-inventor 2R,6R-hydroxynorketamine licensed by Perception Neuroscience → acquired by ATAI) with Sanjay Mathew + Maurizio Fava + Anantha Shekhar J Psychopharmacol 2026-07-13 (DOI 10.1177/02698811261456206, PMID 42438931) "Enhancing plasticity to treat depression and other central nervous system diseases using event-driven pharmacology" — proposes EDP as design framework for plastogen-class CNS drugs (ketamine + classical psychedelics + non-hallucinogenic neuroplastogens) distinct from occupancy-driven pharmacology, argues traditional dose-optimization is counterproductive because it misses the underlying pharmacology; UMB Neuropharmacology thought-leadership positioning for the plastogen class, design-space read on Lieber neuropsych NewCo second-and-third-asset conversation (GPR52 is chronic-agonist occupancy-driven, complementary rather than same class)

Thu, 16 Jul 2026 12:00:00 +0000

Sourcing Radar for Thursday, July 16, 2026. UMB-anchored three-picker — cardio-reperfusion methods-of-use IP filing as direct sourcing lead, cross-institution imaging platform as licensable chassis, and neuropharmacology-framework paper as design-space read on the neuropsych NewCo. **Pick 1 (LEAD, direct sourcing lead) — UMB Anesthesiology + Shock Trauma TLR7/ITGAM cardiac ischemia-reperfusion story with a filed PCT on methods of use.** Li Y (first), Yang Y, Park C, Ren B, Shetty A (UMB Institute for Genome Sciences), Williams B, Zou L, Chao W (senior corresponding; wchao@som.umaryland.edu). "TLR7 Inhibition Limits Ischemic Cardiac Injury by Disrupting ITGAM-Dependent Immune-Endothelial Interactions," *JACC: Basic to Translational Science* 11(8):101626, published online 2026-07-14, DOI 10.1016/j.jacbts.2026.101626, PMID 42447539. All UMB-corresponding from the Translational Research Program in the Department of Anesthesiology and Center for Shock, Trauma and Anesthesiology Research at the University of Maryland School of Medicine. **Mechanism.** Mouse coronary-artery-ligation IR model, 45 min ischemia + 24 h reperfusion. TLR7 KO and pharmacological inhibition with enpatoran (Merck KGaA-developed selective TLR7/8 antagonist in autoimmune clinical trials — SLE, CLE, myositis) both cut MI/AAR ratio significantly and non-additively — epistasis confirms target. Single-nucleus RNA-seq localizes the mechanism to a TLR7-driven inflammatory endothelial program + coordinated myeloid program, both converging on ITGAM (CD11b, α_M integrin) upregulation. Anti-ITGAM antibody (Bio X Cell BE0007) phenocopies TLR7 KO — significantly smaller infarct + markedly reduced neutrophil + monocyte infiltration + epistasis with TLR7 loss. **Effect sizes.** Enpatoran 5 mg/kg intraperitoneal reduces MI/AAR by 35-76% across four dosing windows: prevention (1 h pre-ischemia), treatment (35 min during ischemia), early rescue (30 min post-ischemia), and late rescue (2 h post-ischemia — where efficacy starts to fall off). Peri-procedural PCI window for stent placement is the clinical use-case. **IP disclosure (verbatim).** "A patent application related to use of TLR7/8 antagonist for treatment of cardiac ischemia–reperfusion injury has been filed (PCT/US2025/037133, 'Methods of Treating Ischemia/Reperfusion Injury'). The authors have reported that they have no relationships relevant to the contents of this paper to disclose." **UMB owns the methods-of-use IP; enpatoran composition-of-matter is Merck KGaA's.** Cross-licensing leverage — whoever wants to develop enpatoran into cardio-IR indications needs UMB, and UMB can shop the methods-of-use claim to any TLR7/8 inhibitor competitor program broadly. Funding: NIH R35-GM140822 (Chao); AHA 26POST1559348 (Li). **So what for Blackbird.** Direct sourcing lead on the UM Ventures × Blackbird BioVentures co-investment surface — same track that produced aSKY. UMB SoM methods-of-use IP on a cardio-reperfusion indication where Merck KGaA is de-risking enpatoran tox in autoimmune indications. Two next steps: (1) whether UM Ventures has an active licensing conversation with Merck KGaA on cardio-reperfusion, and (2) whether UMB is positioned to shop the methods-of-use claim to a cardio strategic (Bristol Myers Squibb, Novartis Cardiovascular, CSL Behring — all have interventional-cardiology franchises where a peri-procedural anti-inflammatory adjunct slots cleanly). Action for Hemaka + Eddie — flag Wei Chao + Center for Shock Trauma and Anesthesiology Research for a UM Ventures conversation. R35 (Chao) = maximum-security academic funding + Center for Shock Trauma = deepest translational-anesthesiology infrastructure in the US. Diligence questions: (a) TLR7 blockade impact on acute viral response in peri-procedural window (short duration probably fine, needs characterization); (b) whether MoU claims cover competitor TLR7-only / TLR7/8 molecules broadly or are narrowed to enpatoran; (c) large-animal or human ex-vivo replication of the CD11b-integrin axis before Phase 2. **Pick 2 (platform + landscape) — MIPOT: UMB Diagnostic Radiology cross Johns Hopkins BME cross Kennedy Krieger Kirby Center multimodal imaging chassis for tracking cortical organoid transplants in stroke.** Wang J (first), Qiao G, Tan H, Russel C, Yang B, Li MJ (Princeton ECE), Wang K (JHU BME + Kennedy Krieger F.M. Kirby Center), Xu J (Kennedy Krieger Kirby Center + JHU School of Medicine Russell H. Morgan Department of Radiology and Radiological Science), Chu C, Janowski M, Fu T-M (Princeton), Walczak P (co-supervisor; UMB Diagnostic Radiology), Liang Y (senior corresponding; yajie.liang@som.umaryland.edu). "Intravital Multimodal Imaging of Human Cortical Organoid Transplantation in a Mouse Model of Chronic Stroke," *Adv Sci* 2026, DOI 10.1002/advs.202515913, PMID 42447126, PMC13367743 (open access). **Platform.** MIPOT integrates four modalities on one animal — 9.4 T MRI (T2w + DTI/ADC/FA), bioluminescence imaging via Luc2-mCherry dual reporter driven by human ubiquitin C promoter, intraoperative surgical microscopy, and intravital two-photon fluorescence microscopy using multicolor viral labels (cytosolic mCherry/TagBFP/iRFP-682 + nuclear H2B-GCaMP6s). Novelty is the *combination* plus a two-component 3D-printed removable head-fixation adapter (STL files in supplement) that cycles the same animal through MRI + BLI + TPFM sequentially. **Experimental design + results.** Rose-Bengal photothrombotic stroke in 3-4 mo male C57BL/6J at motor cortex; 10-day infarct maturation; hiPSC (KOLF2.1J) cortical organoid at day 56 differentiation quartered + injected into cleared cavity; cyclosporin A immunosuppression. BLI signal decay: ~50% at D4, ~25% at D7, ~17% at D10, ~5% at D14 (n=15). Endpoint histology at 2 + 4 weeks confirms limited human-graft persistence. **No behavioral efficacy data.** Authors explicitly frame as platform for tracking, not therapeutic candidate. COI: "The authors declare no conflicts of interest." **So what for Blackbird.** Two beats. (1) Licensable imaging chassis — MIPOT sits alongside Kalhor 3DEEP (§2b Kalhor lab spatial-omics platform) as cross-institution Baltimore imaging tools Blackbird could underwrite. Explicit generalization in the discussion: "the MIPOT framework should be broadly adaptable to other in vivo settings involving organoid transplantation" — extends to any CNS cell-therapy diligence. (2) UM Ventures × Blackbird ecosystem read — this paper crosses UMB Diagnostic Radiology (Walczak + Liang) with JHU BME (K. Wang) with Kennedy Krieger Kirby Center (Xu) in a single manuscript, exact operating pattern the UM Ventures × Blackbird co-investment surface is meant to seed. Walczak + Liang are the UMB translational-imaging bench Blackbird should track; this puts them on the map with a real product. Reality-check read on CNS cell-therapy sector — 25% persistence at 2 weeks is a candid limitation that says every group graft-transplanting into infarct cavities needs longitudinal imaging like this. Diligence tool value. **Pick 3 (short landscape/framing, not a sourcing lead) — Gould lab EDP framework for CNS plastogens (UMB Neuropharmacology + Drug Discovery).** Gould TD (first; UMB SoM Departments of Psychiatry, Pharmacology-Physiology and Drug Discovery, and Neurobiology + Veterans Affairs Maryland Health Care System), Mathew SJ (Texas A&M College of Medicine + Baylor + VA), Fava M (MGH Psychiatry), Shekhar A (Pitt SoM). "Enhancing plasticity to treat depression and other central nervous system diseases using event-driven pharmacology," *J Psychopharmacol* 2026, DOI 10.1177/02698811261456206, PMID 42438931. **Framework.** Proposes event-driven pharmacology (EDP) as design framework distinct from occupancy-driven pharmacology, for the class of CNS drugs where a transient binding event produces sustained neuroplasticity that outlasts drug exposure — ketamine as paradigm case, classical psychedelics as extension, non-hallucinogenic neuroplastogens as design target. Argues that developing rapid-acting plastogens with the traditional receptor-occupancy dose-finding model is counterproductive — you optimize for stable receptor occupancy over time and paradoxically get few durable benefits and more side effects, because you miss the underlying pharmacology. **So what for Blackbird — landscape signal only.** Todd Gould is the UMB SoM faculty member who co-invented (2R,6R)-hydroxynorketamine — licensed to Perception Neuroscience, subsequently acquired by ATAI Life Sciences. UMB owns royalties on the estate. Gould publishing a framework paper on how to design + evaluate this class shapes the diligence conversation on every ketamine-adjacent + psychedelic-adjacent asset the venture community sees for the next two years. Design-space read for Lieber neuropsych NewCo second-and-third-asset conversation: GPR52 lead is chronic-agonist occupancy-driven by design (three-symptom-domain thesis), which is correct for the mechanism. Pipeline expansion question is whether the neuropsych NewCo's second + third shots should sit in the plastogen/EDP class rather than the chronic-agonist class. Not a sourcing lead today. Design-space input for Hemaka + Third Rock conversation on NewCo pipeline shape. Full paper links and PCT application number in the show notes.

Sourcing Radar — pipeline day with a first-in-class RNA-degrader modality as the lead sourcing item: Pick 1 Wang N/Mao X/Wang J co-corresponding + Ted Dawson senior + Kennedy Krieger's Mingyao Ying paper in PNAS 2026-07-14 (DOI 10.1073/pnas.2526461123, PMID 42446976) "Targeted alpha-synuclein mRNA degradation by PMO-based RNA-degrading chimeras" introducing an entirely new oligonucleotide modality — phosphorodiamidate morpholino conjugated to a small-molecule warhead that recruits R-Nase L to the 5-prime UTR of SNCA messenger-RNA for endogenous cytoplasmic degradation, doubly orthogonal to the incumbent gapmer-A-S-O/R-Nase H field (BIIB101 failed Phase 1 futility; prasinezumab + cinpanemab washed out Phase 2) — screen of 9 chimeras identified lead 4-D1 with R-Nase-L-dependent alpha-synuclein knockdown in HEK293T + humanized SNCA-mouse primary cortical neurons + patient iPSC-derived cortical neurons, blocks patient-derived fibril prion-like seeding, rescues fibril-driven cytotoxicity, in-vivo mRNA lowering in humanized SNCA mouse — provisional application 63/901,296 "Phosphorodiamidate morpholino oligonucleotides and uses thereof" jointly filed by University of Chicago + Johns Hopkins University, explicitly not licensed, no royalties, Parkinson's Foundation + Maryland Stem Cell Research Foundation funded — direct sourcing lead with MSA + DLB follow-on indications; Pick 2 Zhang/Popel + Fertig co-corresponding + Deshpande paper in same PNAS issue (DOI 10.1073/pnas.2525799123, PMID 42446991) "Quantitative calibration of a spatial QSP model identifies fibroblast impact on HCC immunotherapy" delivering an Approximate Bayesian Computation - Sequential Monte Carlo-calibrated agent-based spatial QSP model of hepatocellular carcinoma with fibroblast-mediated CD8 exclusion module + pretreatment spatial biomarker predicting atezolizumab + TKI response in independent cohort + code on GitHub + Zenodo — two Blackbird beats: rare Johns Hopkins BME + UMB IGS/Greenebaum/UM Institute of Health Computing co-corresponding structure showcases the exact cross-institution model the UM Ventures co-investment agreement should be catalyzing, AND Popel discloses first-time public equity in OptaNova Pharma + Terebra Therapeutics — first confirmation of active Popel-founder equity in OptaNova since 2026-07-06 stealth-stealth flag; Pick 3 landscape signal Zivko/Mahairaki + Lyketsos senior corresponding paper in Alzheimer's & Dementia (NY) TRCI 2026-07-12 (DOI 10.1002/trc2.70271, PMID 42444751) "Heterogeneous responses to memantine in Alzheimer's disease: A precision medicine approach using iPSC models" delivering 19-line iPSC-neuron platform (12 AD + 7 CU, Johns Hopkins ADRC + MATC-sourced) with dose-dependent memantine-driven calcium influx (-9.85% mean, range 0 to -39%) + ROS (-26.05% mean, range -4.23% to -72.21%) modulation and substantial across-line heterogeneity mirroring clinical response — Richman Family Precision Medicine Center of Excellence in AD + three Lieber Institute contributors (Farinelli, Ostlund, Das, Maher) — companion-diagnostic template for memantine and next-wave AD drugs (donanemab, lecanemab), extends Lieber footprint into AD precision medicine, natural second-and-third-asset pipeline route for the neuropsych NewCo

Wed, 15 Jul 2026 12:00:00 +0000

Sourcing Radar for Wednesday, July 15, 2026. Pipeline day — Baltimore preprint pipeline back to normal end-of-week cadence; three-item episode with a genuine first-in-class translational-stage RNA modality out of Johns Hopkins Neurology as the lead sourcing item, and two Baltimore-ecosystem reads with direct portfolio implications behind it. **Pick 1 (LEAD, direct sourcing item) — Dawson lab first-in-class PMO-based RNase-L-recruiting RNA-degrading chimera for alpha-synuclein.** Wang N (first; JHU Institute for Cell Engineering + Neurology + U Chicago Genetic Medicine), Hegde S, Tang Z, Liu H, ..., Ying M (Kennedy Krieger), Mao X (co-corresponding; JHU Neurology + INBT + Materials Science), Wang J (co-corresponding; JHU Neurology), Dawson TM (senior; JHU ICE + Neurology + Snyder Neuroscience + Physiology, Pharmacology & Therapeutics + INBT). "Targeted α-synuclein mRNA degradation by PMO-based RNA-degrading chimeras," *PNAS* 123(29):e2526461123, published online 2026-07-14, DOI 10.1073/pnas.2526461123, PMID 42446976. **Modality.** RIBOTAC in phosphorodiamidate-morpholino chemistry — PMO conjugated to a small-molecule warhead that binds and activates RNase L. Morpholino tags the 5'UTR of SNCA mRNA; warhead recruits cytoplasmic RNase L to that site; endogenous nuclease cuts the transcript. Two orthogonal differentiations at once: PMO backbone rather than gapmer ASO, cytoplasmic RNase L recruitment rather than nuclear RNase H. Either alone is a modality story a translational investor takes; together they define a class. **Preclinical package.** Screen of nine chimeras narrowed to lead 4-D1. RNase-L-dependent knockdown of SNCA mRNA + α-synuclein protein in HEK293T (target-engagement control an FDA reviewer will ask for). Effect replicates in humanized-SNCA-mouse primary cortical neurons + patient iPSC-derived cortical neurons. Blocks prion-like seeding by patient-derived α-synuclein fibrils; rescues fibril-driven cytotoxicity. In-vivo α-synuclein mRNA lowering in humanized SNCA mouse. Complete-to-seed sequence. **IP.** Provisional application 63/901,296 "Phosphorodiamidate morpholino oligonucleotides and uses thereof" jointly filed by University of Chicago and Johns Hopkins University; inventors Ning Wang, Xiaobo Mao, Jingxin Wang. Explicitly not licensed. No royalties. JHTV signal that the program is at term-sheet moment. Funding: NIH R35GM147498, Parkinson's Foundation PF-IMP-1045798 + PF-JFA-1933, Maryland Stem Cell Research Foundation 2019-MSCRFD-4292 + 2024-MSCRFD-6394. **Field context.** Every clinical α-synuclein-lowering program to date runs on ASOs / RNase H. Biogen α-syn ASO failed Phase 1 futility. Prasinezumab + cinpanemab as antibodies washed out. α-syn is a decade-long graveyard for direct-binding modalities against an intrinsically disordered target — orthogonal modalities have real optionality value. **So what for Blackbird.** Exact fit for what the JHU Life Sciences Research Initiative will backfill over the next two years — but composition-of-matter is filed today, sitting at JHTV, unlicensed. Foundational Parkinson's franchise with MSA + DLB as follow-on indications. Founding cast on the paper: Dawson (foundational PD figure) + Mao (α-syn transmission from INBT/Materials Science) + Jingxin Wang (medchem) + Mingyao Ying (Kennedy Krieger). Diligence questions in three places: (1) whether target-conditional RNase L activation avoids interferon-response side effects of global RNase L activation — theory is on-target only; chronic-dosing innate-immune readouts need to be pulled from supplement; (2) CNS delivery of PMO — Sarepta's PMO chemistry is IV/subcut for muscle, CNS delivery is not their franchise, Ionis has ASO-to-CNS know-how competitors do not; (3) in-vivo route, dose, magnitude, biodistribution in humanized-SNCA mouse. Direct action — Hemaka or Avi, flag Ted Dawson's group for a JHTV conversation this week. **Pick 2 (portfolio-adjacent double header) — Popel-Fertig spatial QSP HCC virtual-trial platform + first public disclosure of Popel equity in OptaNova + Terebra.** Zhang S (first; JHU BME), Wang H, Cho Y, Rocha HL (Indiana U), Wong W, Yarchoan M, Jaffee EM, Ho WJ, Kagohara LT (JHU SKCCC + Convergence Institute + Bloomberg-Kimmel), Fertig EJ (co-corresponding; UMB SoM IGS + Greenebaum + UM Institute of Health Computing + Medicine), Popel AS (co-corresponding; JHU BME + Oncology, SKCCC), Deshpande A (JHU Oncology + Convergence + Bloomberg-Kimmel + Data Science & AI Institute). "Quantitative calibration of a spatial QSP model identifies fibroblast impact on HCC immunotherapy," *PNAS* 123(29):e2525799123, published online 2026-07-14, DOI 10.1073/pnas.2525799123, PMID 42446991. **Method.** ABC-SMC pipeline evolves a population of parameter sets so the simulated agent-based spQSP tumor architecture matches spatial-transcriptomics-observed cellular neighborhoods in patient tissue. Fibroblast module added; model reproduces fibroblast-mediated CD8 T-cell exclusion; predicts atezolizumab + TKI response in independent cohort; identifies pretreatment spatial + non-spatial biomarkers. Code + calibrated models on GitHub + Zenodo. Funded by NIH U24CA284156 + U01CA253403 + U01CA212007, DoD, Maryland Cancer Moonshot, Maryland Cigarette Restitution. **Two Blackbird beats.** (1) Platform. Virtual-clinical-trial infrastructure with pretreatment spatial-transcriptomics biomarker readouts is a live licensable asset. Genentech, AstraZeneca, and BMS all run internal QSP groups but none have spatial-omics-calibrated versions in production. Popel lab has been publishing this direction for a decade — this is the maturation moment for a computational-oncology NewCo conversation, code deposited on GitHub lowers friction for licensing dialog. (2) Popel discloses equity in **OptaNova Pharma + Terebra Therapeutics** — first public disclosure of active Popel equity in OptaNova since we flagged it 2026-07-06 as stealth-stealth (Delaware LLC filing only). Confirms Wilmer + JHU-BME peptide-therapeutic cluster (Popel + Green + Pandey + Mirando) as an active-founder cluster with cross-institution academic momentum. Cross-references to a new computational-oncology axis with Fertig at UMB + Deshpande at JHU. **So what for Blackbird.** Case-in-point for the UM Ventures × Blackbird co-investment agreement picking up cross-institution spinout pipeline more actively. Popel + Fertig are the founding cast for a computational-oncology virtual-trial NewCo between JHU BME and UMB IHC. Eddie — has JHTV or UM Ventures had that conversation yet, and if not, this week is the moment. Also confirms OptaNova as a real entity for §2b tracking. **Pick 3 (landscape signal, not a fresh sourcing lead) — Richman Center + Lieber Institute iPSC memantine responder-stratification platform.** Zivko C (first; JHU Genetic Medicine + Richman Family Precision Medicine Center of Excellence in AD), Sagar R, Ahmed W, Xydia A, Farinelli F (Lieber Institute), Ostlund I (Lieber), Das D (Lieber), Maher BJ (Lieber + JHU Psychiatry + Snyder Neuroscience), Lyketsos CG (senior; Richman Center + JHU Psychiatry + Bayview + JHU AD Research Center), Mahairaki V (senior corresponding; JHU Genetic Medicine + Richman Center). "Heterogeneous responses to memantine in Alzheimer's disease: A precision medicine approach using iPSC models," *Alz & Dem (NY) TRCI* 12(3):e70271, published online 2026-07-12, DOI 10.1002/trc2.70271, PMID 42444751, PMC13357697 (open access). **Data.** 19 patient-derived iPSC lines (12 AD, 7 CU — sourced through JHU ADRC + Memory and Alzheimer's Treatment Center + S-CitAD trial) differentiated to NgN2 excitatory cortical neurons. Two functional assays: fluorescence-based calcium influx (Fluo-4 AM, glutamate/glycine stimulation) and oxidative stress (CellROX, menadione stimulation). 24hr memantine pre-treatment: calcium influx -9.85% mean (range 0 to -39%); ROS -26.05% mean (range -4.23% to -72.21%). Across-line heterogeneity dominates individual assay effects. AD lines show higher baseline calcium fluorescence + steeper ROS response slopes. Exploratory / hypothesis-generating; N-power for correlation stratification >1000/group. Lyketsos discloses consulting to Roche, Karuna, Axsome, Otsuka, IntraCellular, BMS, Merck, and roughly a half-dozen more. **So what for Blackbird.** Companion-diagnostic template for memantine (generic, ~1/3 responder rate, $50/mo cost) — the same iPSC-neuron functional-assay platform extends to donanemab, lecanemab, and next-wave symptomatic + disease-modifying AD drugs at meaningful per-patient cost. Richman Center sits inside the same JHU ADRC that Lyketsos runs — the natural home for iPSC-based patient stratification for AD. Lieber contributor list (Farinelli, Ostlund, Das, Maher) extends Lieber footprint from neuropsychiatry into AD precision medicine — natural pipeline expansion route for the GPR52-anchored Lieber neuropsych NewCo when it gets to a second and third asset. Not a direct sourcing lead today; landscape signal + watchlist for how the Richman Center opens to industry partnering. **Editorial framing.** Pipeline refill day after two consecutive lean days. Genuine first-in-class sourcing lead (Dawson RNase-L-recruiting PMO chimera) — direct action for JHTV conversation. Portfolio-adjacent PNAS double delivering both platform-licensing (Popel-Fertig virtual-trial) and OptaNova-cluster confirmation. Landscape signal on Richman Center iPSC precision-medicine platform for AD. Paper links: https://www.pnas.org/doi/10.1073/pnas.2526461123 ; https://www.pnas.org/doi/10.1073/pnas.2525799123 ; https://doi.org/10.1002/trc2.70271

Sourcing Radar — second consecutive lean day; two Johns Hopkins-anchored peer-reviewed picks worth the file: Pick 1 (LEAD, methodology + scouting signal) Sternke/Barrick Nature Chemical Biology 2026-07-10 (DOI 10.1038/s41589-026-02270-6, PMID 42432345) "Protein stability is determined by single-site bias rather than pairwise covariance" — Potts-model isolation of single-site vs pairwise-covariance contributions to protein sequence design demonstrates a genuine stability/activity trade-off (homeodomain ΔG ≈ −14 kcal/mol H-optimized vs −8.2 kcal/mol HJ-optimized; enzyme k_cat higher in HJ-optimized batches; homeodomain DNA-binding Kd 8.9 nM H-optimized vs 343 nM HJ-optimized) — diligence discipline point for the AI-first protein-design pitch cohort Blackbird sees regularly (Cyrus/Xaira/Chai/EvolvePro/Isomorphic Labs) + design-rule for engineered effector proteins on the Aletira SELEXON adjacency + first-author Sternke JHU-Biophysics-to-GSK-Protein-Design-and-Informatics dual affiliation is a scouting signal on where Johns Hopkins protein-engineering IP is flowing; Pick 2 (portfolio-adjacent framing) Kwon/Martinowich Neuropsychopharmacology Hot Topics commentary 2026-07-10 (DOI 10.1038/s41386-026-02497-w, PMID 42432160) "Neuronal genetic risk within the multicellular landscape of schizophrenia" — Lieber-CSO synthesis argues neuronal-genetic-risk populations and broader multicellular tissue-state alterations (astrocytes + oligodendrocytes + microglia + endothelial cells + perineuronal nets + neurovasculature) are complementary rather than contradictory, integrated via shared developmental programs + cell-cell interactions + environmental input + adaptive responses; direct portfolio-adjacency read for the Blackbird-Lieber-Third Rock GPR52 NewCo — soft warning that neuronal-target-alone framing may be insufficient + explicit specification that biomarker/PD strategy for the DC declaration (expected later 2026) should pair GPR52 neuronal-engagement biomarker with a multicellular tissue-state readout (spatial transcriptomics on dosed Lieber postmortem cohort via Martinowich-Maynard-Hicks matched-donor atlas infrastructure, or an equivalent imaging correlate)

Tue, 14 Jul 2026 12:00:00 +0000

Sourcing Radar for Tuesday, July 14, 2026. Second consecutive lean day: 2026-07-11..14 bioRxiv window returned exactly one JHU-corresponding preprint (Bishai TB immunology — no discrete asset), medRxiv returned zero. Two Johns Hopkins-anchored peer-reviewed picks worth pulling for the file. **Pick 1 (LEAD, methodology + scouting signal) — Sternke/Barrick Nature Chemical Biology on single-site vs pairwise-covariance protein-design stability/activity trade-off.** Sternke M (first; dual JHU T.C. Jenkins Department of Biophysics + GSK Protein Design and Informatics, Collegeville PA), Tripp KW, Behera SP, ..., Barrick D (senior corresponding; T.C. Jenkins Biophysics, Johns Hopkins University; barrick@jhu.edu). "Protein stability is determined by single-site bias rather than pairwise covariance," *Nature Chemical Biology*, published online 2026-07-10, DOI 10.1038/s41589-026-02270-6, PMID 42432345, peer-reviewed. **Setup.** Barrick uses Potts models (the sequence-covariance model class underpinning most modern AI-first protein design) to build synthetic sequences that isolate two ingredients — single-site amino-acid preferences (position-level conservation across the evolutionary alignment) vs pairwise correlations between positions (co-evolutionary co-variation). Most current design tools optimize both jointly. Barrick separates them: one batch H-optimized (single-site only), one batch HJ-optimized (single-site + pair correlations), and measures both stability + enzyme activity. **Counter-intuitive finding.** Stability is dominated by single-site bias (homeodomain folding ΔG ≈ −14 kcal/mol H-optimized vs −8.2 kcal/mol HJ-optimized; same pattern across adenylate kinase + DHFR + phosphoglycerate kinase). Activity is dominated by pair correlations (enzyme k_cat substantially higher in HJ-optimized batches; homeodomain DNA-binding Kd 8.9 nM H-optimized vs 343 nM HJ-optimized — ~40-fold affinity difference). Explicit stability/activity trade-off — stability + activity live in different features of the sequence. **So what for Blackbird — two angles.** (1) Diligence discipline point for the AI-first protein-design pitch cohort Blackbird sees regularly (Cyrus Bio, Xaira, Chai Discovery, EvolvePro, Isomorphic Labs). Platform claims about designing thermostable-and-active proteins in one shot should be pressed on which objective the model actually optimizes — if the pitch says stable and active together, the Barrick result names that as a trade-off not a free lunch. Blackbird's diligence question (Anthony/Hemaka/Jonathan/Virginia/Matt Lawler): does the platform quantify the trade-off explicitly, or hide it under a joint objective? (2) Design-rule for the portfolio. Aletira SELEXON platform stacks alternative-splicing modules on AAV; today's payloads are gene-therapy transgenes, but any move into engineered effector proteins (more selective recombinases, Cas variants for gene editing) will confront the same stability-vs-activity trade. Applies equally to any future engineered-enzyme or antibody-scaffold portfolio play. **Scouting signal.** Sternke's JHU Biophysics → GSK Protein Design and Informatics transition is where Johns Hopkins protein-engineering talent is flowing. Esther + Avi action: quiet check with JHTV biophysics-side portfolio manager on whether any Barrick-lab IP has not already been optioned into a GSK relationship. (No COI or patent-application declaration surfaced through the r.jina.ai proxy on the Nature Chemical Biology article page; peer-reviewed methodology with GitHub code + MSA release, consistent with open academic disclosure rather than filed provisional — but still worth the JHTV check.) **Pick 2 (portfolio-adjacent framing) — Kwon/Martinowich Lieber Institute Neuropsychopharmacology Hot Topics commentary; biomarker/PD spec sharpener for GPR52 NewCo.** Kwon SH (first) + Martinowich K (senior corresponding), "Neuronal genetic risk within the multicellular landscape of schizophrenia," *Neuropsychopharmacology* Hot Topics, published online 2026-07-10, DOI 10.1038/s41386-026-02497-w, PMID 42432160. Both authors Lieber Institute for Brain Development; Martinowich additionally Solomon H. Snyder Department of Neuroscience + JHU Psychiatry and Behavioral Sciences + JHU Kavli Neuroscience Discovery Institute. Competing interests: none declared. **Argument.** Schizophrenia genetics (PGC wave-3 + Pergola trans-eQTL 641-novel-gene work) points to neuronal populations as primary carriers of genetic risk, but the postmortem molecular literature (including Lieber's own spatial-transcriptomics program) keeps finding alterations across a much broader multicellular landscape — astrocytes, oligodendrocytes, microglia, endothelial cells, perineuronal nets, neurovasculature. Kwon + Martinowich argue these are complementary not contradictory, integrated via shared developmental programs, cell-cell interactions, environmental input, adaptive responses. Programmatic question they leave the field on: how does neuronal vulnerability become wired into a broader multicellular pathobiology. **So what for Blackbird — GPR52 program biomarker/PD framing.** GPR52 is a neuronally-expressed orphan GPCR; the Blackbird-Lieber-Third Rock development thesis is that neuronal GPR52 agonism engages all three symptom domains (positive, negative, cognitive). Martinowich commentary is a soft warning (in Lieber's own voice) that neuronal-target-alone framing may not be sufficient — if the schizophrenia phenotype is a stable multicellular state including glial + vascular contributions, a purely neuronal agonist may not collapse it end-to-end. Does NOT change the GPR52 thesis (downstream circuit engagement should still drive functional outcomes, and Third Rock DC plan already contemplates broad symptom-domain readouts). Does explicitly specify where the biomarker + PD strategy should evolve: pair the GPR52 neuronal-engagement biomarker with a multicellular tissue-state readout — spatial transcriptomics on a dosed Lieber postmortem cohort (Martinowich-Maynard-Hicks matched-donor atlas infrastructure is exactly the substrate) or an equivalent imaging correlate capturing glial + vascular components. Hemaka action: surface directly with Third Rock team ahead of DC declaration expected later 2026 — Lieber CSO framing is what the Third Rock partner will be reading anyway. **Editorial framing.** Second consecutive lean day. Two-pick platform-and-landscape episode is the honest read. Baltimore preprint pipeline was still thin through Monday night; expect refill Wednesday/Thursday on normal end-of-week posting cadence. If Wong caspase-3-tau, Luetkens cathepsin-B, or Kalhor 3DEEP moves IP into JHTV or UM Ventures prosecution, that becomes the lead the moment it hits the file. Portfolio memory updated: Barrick methodology note added to §2b JHU-adjacent ecosystem section; Martinowich commentary link added to §2 GPR52 NewCo entry as biomarker/PD spec-sharpener signal ahead of DC declaration. Paper links: https://www.nature.com/articles/s41589-026-02270-6 ; https://www.nature.com/articles/s41386-026-02497-w

Sourcing Radar — a candid lean Monday after Sunday cleared the strongest translational items (Slusher/Kamiya astrocytic G-C-P-two P-O-C-D + Maqsood/Hoag submucosal-lift hydrogel), so today is a two-pick platform + landscape episode: Pick 1 Chen (Stanford Statistics) + Hicks (JHU Biostatistics + Malone Center; co-developer of spatialLIBD stack + co-author on the Lieber-Hopkins matched-donor dlPFC 2024 + dACC 2026 Visium atlases) bioRxiv preprint 2026-07-09 (DOI 10.64898/2026.07.05.736638) "Mind the alignment gap: a spatial transcriptomics benchmark for scientific coding agents" — 40 SABench spatial-alignment tasks converted into hidden-scored A-I coding-agent workloads run under three scaffolding conditions (Basic bare-bones prompt, Package-Aware prompt naming PASTE + Spateo + SPACEL + STAligner, and Full+prior with pre-built virtual environment) using Codex-C-L-I on gpt-5.3-codex xhigh reasoning; counter-intuitive finding — mean composite alignment score DROPPED from 0.43 (Basic) to 0.36 (FPP) with 95% C-I [-0.11, -0.03] excluding zero, despite specialized-package usage going from ~20% to 72% (PASTE2) as scaffolding richened; trace inspection revealed richer scaffolding induced unnecessary transformations + brittle package-first workflows + infrastructure failures where the agent stopped reasoning from first principles and started plumbing; direct portfolio-adjacency read for Blackbird — Hicks is the same group producing the Lieber-Hopkins atlas platform (Martinowich/Maynard co-authorship on dACC atlas covered 2026-07-10 Radar) and is now embedding agent-eval methodology in the same domain, meaning the counter-intuitive tooling result is a canary Blackbird portfolio cos (Aletira, aSKY, Lieber neuropsych NewCo, 1104health) should internalize before committing to agent-scaffolding investments + calibrates further Kalhor 3-D-E-E-P deep-tissue chassis cluster underwriting rationale; Pick 2 Lim (Stanford Neurosurgery, ex-JHU Chair) + Grossman/Ye (JHU SKCCC) + Adult Brain Tumor Consortium ABTC 1501/NCT02658981 Nature Medicine 2026-07-10 (DOI 10.1038/s41591-026-04475-7, PMID 42432293) anti-L-A-G-three (relatlimab, B-M-S) ± anti-P-D-one (nivolumab, B-M-S) Phase 1 in recurrent glioblastoma — 46 patients enrolled 2016-2020, MTD 800mg relatlimab mono / 160mg + 240mg nivo combo, 0/23 D-L-T monotherapy vs 6/23 (26%) combo, 12-month O-S 34.8% mono vs 52.2% combo (descriptive, non-randomized), no confirmed R-A-N-O responses (25 pseudoprogression), 11% long-term survivors past 24 mo correlated with baseline type-I I-F-N signature + T-cell clonality + myeloid A-P-C; interpret as modest ambiguous biomarker-consistent checkpoint signal in a graveyard indication (Checkmate-143 nivo mono failed + K-EYNOTE + REGN2810 all disappointing) meaning the case for a mechanistically-orthogonal metabolic angle to G-B-M (Paul Watkins A-C-S-V-L-three grassofermata sourcing lead from 2026-07-08) gets a little stronger and the "no new drug in 20 years" narrative gets more concrete — landscape-level input to the Watkins conversation, not a new sourcing lead

Mon, 13 Jul 2026 12:00:00 +0000

Sourcing Radar for Monday, July 13, 2026. Candid lean-day framing: Sunday cleared the strongest translational items (Slusher/Kamiya astrocytic GCPII POCD indication-expansion + Maqsood/Hoag UMSOP+JHU-GI submucosal-lift hydrogel), and Monday morning does not have a same-caliber sourcing lead. Two-pick episode: one JHU-primary platform read that touches the Lieber-Hopkins spatial-omics arc, and one landscape context piece that sharpens an existing Blackbird sourcing thesis (Watkins ACSVL3 GBM metabolic vulnerability from 2026-07-08). **Pick 1 (LEAD, platform read) — Chen/Hicks spatial-transcriptomics AI-agent benchmark; canary for portfolio agent-scaffolding investments.** Chen YT (Stanford Statistics; first) + Hicks SC (senior corresponding; JHU Bloomberg SPH Biostatistics + JHU Malone Center for Engineering in Healthcare). "Mind the alignment gap: a spatial transcriptomics benchmark for scientific coding agents," bioRxiv v1 2026.07.05.736638, posted 2026-07-09, DOI 10.64898/2026.07.05.736638. **Setup.** 40 spatial-alignment tasks stratified from SABench (Zeira et al. computational benchmark on aligning 2D tissue slices via coordinate maps between pairs of transcriptomics measurements) converted into hidden-scored executable workloads. Same agent (Codex-CLI on gpt-5.3-codex; reasoning=xhigh; approval-bypass), same tasks, three configurations of increasing scaffolding richness: Basic bare-bones prompt, Package-Aware prompt naming PASTE + Spateo + SPACEL + STAligner, Full+Prior (FPP) with pre-built virtual environment where those packages already installed. Hidden scorer + leakage audit + 3 independent replicates per (task, config). **Counter-intuitive finding.** Mean composite alignment score dropped from 0.43 (Basic) to 0.36 (FPP) with 95% CI [-0.11, -0.03] excluding zero — richer scaffolding hurt performance despite package-use going from ~20% PASTE (Basic) to 72% PASTE2 + 68% PASTE + 54% Spateo + 43% SPACEL (FPP). Trace inspection: added tooling induced unnecessary coordinate transformations, brittle package-first workflows (agent stopped reasoning from first principles and started plumbing), and infrastructure failures where an installed package misbehaved and the agent did not recover. **So what for Blackbird — two reasons this matters.** (1) Portfolio-adjacency signal. Hicks is co-senior on the Lieber-Hopkins matched-donor Visium atlas program (Martinowich/Maynard co-authorship on 2024 dlPFC atlas + 2026 dACC atlas covered in 2026-07-10 Radar) and co-developer of the spatialLIBD stack that a lot of academic + industry spatial-transcriptomics analysis runs on. The same lab producing the atlas methodology is now embedding agent-eval methodology in the same domain — meaning the tooling result is a canary Blackbird portfolio cos (Aletira, aSKY, Lieber neuropsych NewCo, 1104health) should internalize before committing to agent-scaffolding investments. Diligence discipline: inspect agent traces, not just final output scores. (2) Kalhor 3DEEP deep-tissue chassis underwriting rationale. Strengthens the cluster case — atlases (Lieber) + deep-tissue platform (Kalhor) + agent-eval methodology (Hicks) all clustering in the same Hopkins-Wilmer-BME-Lieber orbit. **Actions.** Circulate paper to portfolio-co CTO track. Hemaka — is there an active conversation channel to Hicks lab that could carry an agent-scaffolding-discipline conversation into portfolio cos? Avi — pull for Kalhor 3DEEP diligence file as adjacent-methodology validator. **Pick 2 (landscape context) — Lim/Grossman anti-LAG-3 ± anti-PD-1 Phase 1 in recurrent GBM; sharpens the Watkins ACSVL3 sourcing thesis.** Lim M (senior; Stanford Neurosurgery, ex-JHU Chair) + Ye X + Grossman SA (JHU SKCCC), "Anti-LAG-3 with or without anti-PD-1 in recurrent glioblastoma: a phase 1 trial," *Nature Medicine*, published online 2026-07-10, DOI 10.1038/s41591-026-04475-7, PMID 42432293. Adult Brain Tumor Consortium ABTC 1501 / NCT02658981. **Design + data.** 46 patients recurrent GBM enrolled 2016-2020. Relatlimab (BMS anti-LAG-3) mono up to 800mg vs relatlimab 80-160mg + nivolumab 240mg combo (23 per arm). Neoadjuvant pre-surgical dose component. **Safety.** 0/23 DLTs mono; 6/23 (26%) DLTs combo (cerebral edema, hypertension, thyroiditis). MTD 800mg mono / 160mg+240mg combo. **Efficacy (descriptive, non-randomized).** 12-mo OS 34.8% mono vs 52.2% combo. No confirmed RANO responses; 25/46 pseudoprogression. 5/46 (11%) long-term survivors past 24 months. Favorable-outcome baseline signatures: elevated type-I/II IFN signaling, higher T-cell clonality, more myeloid APC. Neoadjuvant dose associated with increased intratumoral CD8+ T-cell infiltration. **Interpretation.** Recurrent GBM has been the graveyard of checkpoint programs for a decade — Checkmate-143 nivo mono failed; pembrolizumab + cemiplimab similarly disappointing. LAG-3 engagement produces a modest, ambiguous, biomarker-consistent signal in a small Phase 1 that BMS can either move to Phase 2 or park. Not a clean win, not a clean loss. **So what for Blackbird.** Landscape input to the Paul Watkins ACSVL3/grassofermata sourcing thesis carried in memory from 2026-07-08 Radar — every ambiguous checkpoint-combination readout in GBM makes the case for a mechanistically-orthogonal metabolic angle a little stronger, and the "no new drug in 20 years" narrative more concrete for a translational investor pitch. Not a new item; a same-week peer-reviewed context that changes how a Watkins diligence meeting reads. Hemaka/Avi — worth pulling for the file if actively tracking Watkins angle. Grossman + Ye = active JHU-side ABTC coordination; potential channel for interfacing with the JHU GBM neurosurgical/neuro-oncology community that a metabolic-vulnerability program will need to socialize into. **Editorial framing.** Honest lean-Monday two-pick episode. No sourceable partner-institution lead in the 3-4 day window; Sunday cleared the strongest translational items. Today's value is platform-calibration (Hicks agent-scaffolding canary — actionable read for Blackbird portfolio co CTOs across the board) + landscape-context (Lim GBM Phase 1 sharpens Watkins ACSVL3 diligence framing). Baltimore preprint funnel expected to refill Tuesday/Wednesday on the normal end-of-week cadence; several marquee items being tracked (Wong caspase-3-tau, Luetkens cathepsin-B, Kalhor 3DEEP) will surface as leads the moment IP moves. Paper links: https://www.biorxiv.org/content/10.64898/2026.07.05.736638v1 ; https://www.nature.com/articles/s41591-026-04475-7

Sourcing Radar — a Sunday two-episode day with the strongest indication-expansion signal of the week for Blackbird's existing Slusher G-C-P-two I-B-D program: Hasegawa/Kamiya senior corresponding + Barbara Slusher senior middle-author paper in Alzheimer's & Dementia July 2026 (DOI 10.1002/alz.71666, PMID 42427142) delivers a mechanistically clean, sex-stratified, druggable G-C-P-two node for perioperative cognitive dysfunction — aged C57BL/6 male mice undergoing abdominal surgery develop recognition + spatial memory deficits, hippocampal astrocyte scRNA-seq identifies a G-C-P-two-upregulating subpopulation, AQP4 polarization at astrocytic endfeet collapses (glymphatic impairment), astrocyte-specific G-C-P-two knockdown and Slusher's own 2-P-M-P-A both rescue cognition + aquaporin-4 polarization + hippocampal glutamate levels, no competing interests declared, NIH NIA/NIDA/NIMH funded — meaning JHU Drug Discovery's C-N-S-penetrant G-C-P-two chemotype (built alongside the gut-restricted I-B-D program) has a natural second-indication target with sex-biomarker-selectable Phase 2 design in a P-O-C-D unmet need with no approved therapy across ~10-20M/yr US elderly surgeries; plus Pick 2 Maqsood/Hoag senior corresponding UMSOP + JHU GI (Khashab + Ngamruengphong) + JHU Radiology (Krimins) sodium alginate submucosal lifting hydrogel with integrated methylene blue in Advanced Healthcare Materials 2026-07-11 (DOI 10.1002/adhm.71427, PMID 42435344), Maryland Innovation Initiative funded — a real device-shaped commercial improvement (viscosity 44-190 mPa∙s vs commercial 1.7-44; ex vivo 60min lift + in vivo swine 45min cushion retention) over Boston Scientific ORISE Gel + Eleview + EverLift + BlueBoost + EndoClot SIS on the UM Ventures × Blackbird co-investment surface (aSKY track); plus Pick 3 landscape-note Prigge JHU Bloomberg SPH Feinstone Dept Molecular Microbiology & Immunology + Sigala (Utah, ex-JHU) co-senior bioRxiv 10.64898/2026.03.29.713301 identifying a first-in-class antimalarial target hub — apicoplast acyl carrier protein essential in blood-stage P. falciparum via FASII-independent role stabilizing apicoplast pyruvate kinase II (sole NTP source; loss collapses isoprenoid synthesis + apicoplast biogenesis) — flagged as target biology worth the file but lower-margin for Blackbird's thesis

Sun, 12 Jul 2026 12:00:00 +0000

Sourcing Radar for Sunday, July 12, 2026. Two-episode day — Portfolio Watch running in parallel. **Pick 1 (LEAD) — Slusher/Kamiya astrocytic GCPII as druggable node for perioperative cognitive dysfunction.** Hasegawa Y (first), Kawai H, Wiseman R, Ursini G (Lieber), Alton H, Xiong F (Mount Sinai), Gummadi S, Mao N, Mailhot K, Li Y, Zhu X, Sawada T (Lieber), Slusher B (JHU Drug Discovery + Pharmacology & Molecular Sciences; senior middle), Kamiya A (senior corresponding). "Glutamate carboxypeptidase II activation in astrocytes mediates glymphatic impairment and cognitive vulnerability in the aging brain following surgery," *Alzheimer's & Dementia* 22(7):e71666, 2026 (DOI 10.1002/alz.71666, PMID 42427142). **Findings.** Aged C57BL/6 mice + abdominal surgery model of POCD. **Male-specific vulnerability** in recognition + spatial memory. Hippocampal astrocyte scRNA-seq identifies a GCPII-upregulating subpopulation; AQP4 polarization collapses at astrocytic endfeet (glymphatic impairment). Astrocyte-specific GCPII KD + pharmacologic 2-PMPA (Slusher lab compound) both rescue cognition + AQP4 polarization + hippocampal glutamate. No competing interests. NIH NIA/NIDA/NIMH funded. **So what for Blackbird.** Direct indication-expansion signal for the existing JHU Drug Discovery GCPII portfolio. Blackbird already has the Slusher gut-restricted GCPII program for Crohn's/UC — this paper opens a second brain-selective indication (POCD, ~10-20M/yr US elderly surgeries, no approved therapy) using the CNS-penetrant chemotype the Slusher lab already has in-house. Sex-stratified mechanism → biomarker-selectable Phase 2 design (male patients over 65, AQP4 imaging surrogate, peri-op dosing). Resonates with the Huganir sex-differentiated synaptic-protein turnover paper we covered on 2026-07-09 — sex is emerging as a first-class translational-design axis at Hopkins this week. **Actions.** Hemaka + Avi — pull the paper, conversation with Slusher on whether the CNS-penetrant chemotype is ready to spin into an indication-expansion enabling package alongside the existing IBD program. Diligence: CNS-penetrant chemotype status, brain PK + target-engagement biomarkers in aged animals, market sizing all-comers vs biomarker-selected, regulatory path (no formal FDA guidance for POCD endpoints — needs scoped early). Kamiya lab is the mechanism + imaging partner; Slusher has the medchem stack. **Watch.** JHTV posture on caspase-3-tau versus POCD claims within the broader Slusher GCPII estate; whether the C-N-S penetrant chemotype has an active back-up license/option currently. **Pick 2 — Maqsood/Hoag UMSOP + Khashab/Ngamruengphong JHU GI sodium alginate submucosal lifting hydrogel with methylene blue.** Maqsood I (first), Wang Y, Freel K, Ibrahim A, Akshintala VS (JHU GI), Evani ST (JHU GI), Krimins RA (JHU Radiology), Ngamruengphong S (JHU GI), Khashab M (JHU GI), Hoag SW (senior corresponding; UMSOP). "Sodium alginate-based submucosal lifting hydrogel with integrated visualization for endoscopic resection," *Adv Healthc Mater*, DOI 10.1002/adhm.71427, PMID 42435344, published 2026-07-11. Funded by Maryland Innovation Initiative. **Findings.** Shear-thinning Na-alginate + methylene blue hydrogel. Viscosity 44-190 mPa∙s at low shear (vs commercial agents 1.7-44). Lyophilized reconstitution 60-90 s. Physiological pH ~6.5 + osmolality 265-275 mOsm/kg. Ex vivo: sustained lift 60 min. In vivo (swine): cushion retention 45 min, clean injectability, no adverse tissue reactions. **So what for Blackbird.** Real device-shaped commercial improvement over Boston Scientific ORISE Gel (~$150M/yr GI-endoscopy accessory) + Eleview + EverLift + BlueBoost + EndoClot SIS on the UM Ventures × Blackbird co-investment surface (aSKY track). Cross-campus UMSOP + JHUSOM inventorship — UM Ventures + JHTV IP-allocation conversation needed. Not a NewCo candidate; more a license-into-device-portfolio-company or a mid-cap acquirer play. Straightforward 510(k) path. **Actions.** Bridget — pull the paper. Eddie — U-M Ventures posture on composition-of-matter filing + JHTV IP-allocation split. Landscape scan of GI-endoscopy accessory acquirers (Boston Scientific incumbent, Olympus, Fujifilm, ConMed). **Pick 3 — landscape note. Prigge JHU Bloomberg SPH + Sigala (Utah) apicoplast ACP-PKII axis, a first-in-class antimalarial target hub.** Geher SWR (first) + Prigge ST (co-senior; JHU Bloomberg SPH, W. Harry Feinstone Dept of Molecular Microbiology & Immunology) + Sigala PA (Utah, corresponding, ex-JHU). bioRxiv v1, 2026-03-29, DOI 10.64898/2026.03.29.713301 — surfaced in the JHU bioRxiv channel this week. **Findings.** ACP knockout/knockdown lethal to blood-stage P. falciparum via FASII-independent mechanism — ACP physically binds and stabilizes apicoplast pyruvate kinase II (PKII), the sole NTP source in the organelle; loss of PKII collapses isoprenoid synthesis + apicoplast biogenesis. Genuinely new mechanistic target hub. **So what for Blackbird.** Landscape note only — Blackbird's translational thesis lives in higher-margin indications than antimalarials, and Prigge historically moves discoveries through global-health partnerships (MMV, Gates) rather than venture-backed startups. But if a medchem-heavy neglected-disease vehicle comes across BD, this target hub is worth flagging. **Editorial framing.** Sunday double-episode day. Slusher GCPII POCD indication-expansion is the highest-value item — natural pipeline expansion of an existing Blackbird position on a target where JHU Drug Discovery holds the world's deepest medchem stack, in an indication with no approved therapy and a biomarker-selectable Phase 2 design. UMSOP/JHU-GI hydrogel is a real product-shaped device story on the UM Ventures × Blackbird surface. Prigge apicoplast is landscape-level target biology worth the file. Portfolio Watch drops in parallel. Paper links: https://doi.org/10.1002/alz.71666 ; https://doi.org/10.1002/adhm.71427 ; https://doi.org/10.64898/2026.03.29.713301

Portfolio Watch — Nxera Pharma is actively shopping the Phase-2-ready NXE'149 G-P-R fifty-two agonist for schizophrenia post-Boehringer walkaway (Dec 2025), which lands the single most important development for Blackbird's portfolio this week: a direct same-mechanism competitor to the Lieber Institute Neuropsych NewCo may be picked up by a specialty acquirer 12-24 months ahead of Lieber's own DC declaration expected later 2026 — tightening the calendar in the middle of Kynexis KYN-5356 KAT II Q4 2026 Phase 2 CIAS readout (Schwarcz-cofounded, licensed from Mitsubishi Tanabe) and B-M-S Cobenfy ADEPT-2 AD-psychosis reset (six-to-nine-month delay in muscarinic-class share-of-voice on cognition = partial offset); plus F-D-A formally opened the search for a permanent C-B-E-R Office of Therapeutic Products director (applications due July 21, Karim Mikhail overseeing) meaning the review-culture pivot toward rare-disease AAV flexibility hangs on interim leadership through Q3/Q4 while Regeneron Otarmeni maintains its zero-dollar U-S price floor on the OTOF-AAV market (bad for Sensorion monetization; neutral-to-good for Aletira's non-OTOF cell-type-selective positioning); plus Slusher G-C-P-two I-B-D threat landscape crowding fast — Merck tulisokibart ATLAS-UC Phase 3 TL1A win (June 22 spillover; first anti-TL1A Ph3-positive registrational induction data + 3 new Ph2b indications) + Vera TRUTAKNA atacicept accelerated approval July 7 as first dual BAFF+APRIL for IgA nephropathy (5th IgAN drug/4th mechanism in ~2 years) signaling the entire autoimmune indication cluster is being drained fast, meaning the Slusher differentiation must lead on the enteric-neuropathic-pain axis or the positioning is not defensible; plus a real validating datapoint for Blackbird's R-N-A translation-activator platform in Skyhawk Therapeutics' twelve-month interim from SKY-0515 Phase 1/2 in Huntington's (up to 69% mutant huntingtin reduction + cUHDRS slight improvement in >175 patients dosed, oral CNS-penetrant small-molecule RNA splicing modulator now expanding into FALCON-HD Phase 2/3 US sites in July 2026) — the strongest single existence-proof for the small-molecule-modulates-RNA modality this year; plus F-D-A real-time clinical trials pilot (R-T-C-T) selection criteria disseminated July 2026 with AstraZeneca TrAVeRse mantle-cell + Amgen STREAM-SCLC on Paradigm Health infrastructure as a direct tailwind for 1104health's clinician-facing trial-marketplace positioning with competitive pressure from Massive Bio Reticulum Nexus + April OpenAI partnership

Sun, 12 Jul 2026 12:30:00 +0000

Portfolio Watch for the week ending Sunday, July 12, 2026. Sunday two-episode day (Sourcing Radar also dropping today). The single most important development for the portfolio this week is on the competitive side: Nxera shopping the Phase-2-ready GPR52 agonist directly threatens Lieber NewCo positioning ahead of its own DC declaration. **Section 1 — Lieber neuropsych NewCo (GPR52) — competitive landscape.** **Nxera Pharma NXE'149 (ex-Sosei Heptares HTL-0016878) GPR52 agonist actively shopping.** Boehringer Ingelheim optioned Nxera's Phase-2-ready GPR52 agonist in 2024, was expected to take up the option Q3/Q4 2025, and instead walked away in December 2025. Nxera has been shopping the program since. **Only other Phase-2-ready GPR52 asset in industry.** The Lieber NewCo is still working toward a DC declaration expected later 2026. If a specialty neuro buyer (Otsuka, Karuna-into-BMS, Neurocrine, AbbVie-Cerevel) picks up NXE'149 in Q3 2026, Lieber NewCo enters DC declaration 12-24 months behind a direct same-mechanism competitor. **Actions.** Eddie/Hemaka — track bidder set through Q3; make sure Lieber NewCo differentiation story sharpens now (best-in-class GPR52 chemistry vs first-Phase-2). Matt/Hemaka — conversation with Third Rock counterparts on whether Lieber DC timing needs to compress from later 2026 into Q3. Sources: https://www.fiercebiotech.com/biotech/boehringer-walks-away-nxeras-phase-2-ready-schizophrenia-program ; https://www.pharmaceutical-technology.com/news/nxera-seeks-schizophrenia-programme-buyer-after-boehringer-snub/ **Kynexis KYN-5356 KAT II Phase 2 CIAS topline Q4 2026 per company disclosure.** Licensed from Mitsubishi Tanabe, Schwarcz (UMB) is scientific cofounder, Forbion-led €57M Series A Nov 2023. Human PoM data (Schwarcz/Buchanan tryptophan challenge) covered in 2026-07-10 Radar. **Read.** Kynexis reads out on cognition ahead of Lieber DC. If positive, the "all three symptom domains" narrative gets its cognition beat validated by a competitor first — not a threat to the mechanism but a threat to the pole-position framing. Source: https://www.kynexistx.com/news/phase-2-first-patient-dosed **BMS Cobenfy ADEPT-2 AD psychosis reset (threat-softening).** BMS paused enrollment in ADEPT-2 after disclosing enrollment-integrity issues at a subset of sites; trial being extended with additional enrollment; full ADEPT readout that was tracking to earlier this year now stacks into end of 2026. **Read for Lieber NewCo.** Near-term muscarinic-class share-of-voice compression in cognition + indication-expansion narratives just got pushed back 6-9 months. Modest tailwind on launch positioning window; does not resolve Cobenfy already being approved in SCZ. Source: https://www.psychiatrictimes.com/view/delay-on-cobenfy-adept-alzheimer-disease-psychosis-data-following-phase-3-study-irregularities-new-patients-enrolled **Section 2 — Aletira Therapeutics — regulatory and competitive landscape.** **FDA formally opened search for permanent CBER Office of Therapeutic Products director, applications due July 21.** Karim Mikhail (acting CBER director) oversees during interim. Formal search opening = interim reporting arrangement will last at least through Q3, probably into Q4. **Read for Aletira.** Permanent-director gap at OTP risks slowing/destabilizing the review-culture pivot toward more rare-disease AAV flexibility that Vijay Kumar (covered last week) had been steering. Watch-item drag, not existential. Eddie tracking external-candidate short list. Source: https://insights.citeline.com/pink-sheet/agency-leadership/us-fda/us-fda-begins-search-for-cell-gene-therapy-office-director-KBRBLF463VETPP4RC6GYFGZA7A/ **Regeneron Otarmeni (lunsotogene parvec-cwha, ex-DB-OTO) — sole approved OTOF AAV, free-in-US commercial model.** Approved April 2026 under CNPV pilot. Zero-dollar price floor for second-in-class OTOF entrants. **Read for Aletira.** Non-OTOF cell-type-selective positioning well insulated. Aletira does not compete on OTOF; target-white-space read on non-OTOF hereditary hearing loss remains open. No new Sensorion SENS-501 data in the window (last was March 6-month Audiogene). Sources: https://investor.regeneron.com/news-releases/news-release-details/otarmenitm-lunsotogene-parvec-cwha-approved-fda-first-and-only ; https://www.pharmiweb.com/press-release/2026-03-23/sensorion-reports-six-month-update-from-the-audiogene-phase-12-trial-of-sens-501-and-advances-gjb2 **Section 3 — Slusher GCPII IBD — competitive threat landscape.** **Merck tulisokibart ATLAS-UC Phase 3 induction win (June 22 spillover).** First anti-TL1A biologic to hit primary + key secondary in Ph3 registrational induction study in moderately-to-severely active UC. Merck also just initiated 3 new Ph2b indications for tulisokibart. **Read for Slusher GCPII.** IBD competitive space consolidating faster than expected around biologics with prevalent-disease reach (tulisokibart TL1A + Abivax obefazimod oral miR-124 + JNJ icotide ICONIC-UC + JNJ Tremfya vs AbbVie Skyrizi Crohn's H2H). The differentiation story for oral gut-restricted GCPII must lead on the pain axis — bimodal effect on inflammation + enteric neuropathic pain. Source: https://www.merck.com/news/mercks-tulisokibart-met-primary-and-key-secondary-endpoints-in-the-phase-3-atlas-uc-induction-only-study-in-patients-with-moderately-to-severely-active-ulcerative-colitis-uc/ **Vera Therapeutics TRUTAKNA (atacicept-vymj) FDA accelerated approval July 7 — first dual BAFF+APRIL for IgA nephropathy.** 5th IgAN drug / 4th mechanism in ~2 years (Tarpeyo → Filspari → Fabhalta → Voyxact → TRUTAKNA). Foresite-adjacent mechanism-repurposing playbook off a 2010 Merck-Serono lupus asset. **Read for Slusher GCPII IBD (indirect).** Autoimmune indication cluster being drained at unusual pace — IBD bucket already has TL1A + IL-23 + miRNA + integrin-α4β7 staked. GCPII's differentiation must be crisp enough to survive autoimmune-market compression + tighter mechanistic specialization. Source: https://www.globenewswire.com/news-release/2026/07/07/3323532/0/en/vera-therapeutics-receives-fda-accelerated-approval-for-trutakna-for-adult-patients-with-primary-iga-nephropathy.html **Section 4 — RNA translation-activator platform (Blackbird) — validation.** **Skyhawk SKY-0515 twelve-month interim Phase 1/2 in Huntington's disease + FALCON-HD Phase 2/3 US expansion July 2026.** Oral CNS-penetrant small-molecule RNA splicing modulator. Twelve-month data: up to 69% mutant huntingtin reduction, modest clinical improvement on cUHDRS composite, clean safety in >175 patients dosed. **Read for Blackbird translation-activator platform.** Strongest single existence-proof for small-molecule-RNA-modulation category at registrational scale. Blackbird's translation-activator platform (translation activation, not splice modulation) sits adjacent — same "small molecule increases rather than silences target RNA" category thesis. **Actions.** Avi should have the Skyhawk 12-mo package as a citable proof point in the next outside pitch. Sources: https://en.hdbuzz.net/skyhawk-shares-a-closer-look-at-sky-0515-data-plus-a-peek-at-expanding-us-trial-access/ ; https://www.prnewswire.com/news-releases/skyhawk-therapeutics-announces-twelve-month-interim-results-from-phase-12-clinical-trial-of-sky-0515-in-huntingtons-disease-302786696.html **Section 5 — 1104health — regulatory tailwind (brief).** **FDA real-time clinical trials (RTCT) pilot selection criteria disseminated July 2026 with AstraZeneca TrAVeRse (mantle cell lymphoma) + Amgen STREAM-SCLC on Paradigm Health infrastructure.** **Read.** FDA endorsement of tech-mediated real-time trial data review = direct thesis tailwind for 1104health's clinician-facing trial-marketplace positioning. Competitive pressure: Massive Bio Reticulum Nexus + April 2026 OpenAI partnership positioning as "AI operating system for oncology access." **Action.** Rose should be visible to FDA on RTCT selection criteria window + push for pilot inclusion; if Massive Bio wins first hospital-system contract in H2 2026, segment shape locks in. Sources: https://www.fda.gov/news-events/press-announcements/fda-announces-major-steps-implement-real-time-clinical-trials ; https://www.fiercebiotech.com/biotech/fda-unveils-plan-real-time-review-clinical-trial-data-astrazeneca-and-amgen-already-board **Editorial framing.** The week's biggest story = Nxera GPR52 shopping in same quarter as Kynexis Q4 KAT II readout on cognition. Lieber NewCo needs to move faster on DC + sharpen differentiation before a same-mechanism competitor is picked up by a specialty acquirer. Aletira regulatory picture on hold pending permanent OTP director but non-OTOF white-space intact. Slusher GCPII IBD positioning needs to lead on the pain axis — biologics consolidating faster than expected. Skyhawk gave the RNA-modulator category its cleanest 12-mo data point yet, cite on translation-activator side. 1104health has a real FDA-endorsed tailwind via RTCT pilot.

Sourcing Radar — a U-M-B day: three separate labs at the University of Maryland, Baltimore have posted translational work over the past four days that constitutes the strongest single-week U-M Ventures pipeline signal in months — LEAD is Rehman/Giese/Riazuddin/Frolenkov/Ahmed (UMB SoM ENT c-E-A-R Lab senior corresponding zmahmed@som.umaryland.edu) Molecular Therapy 2026-07-09 (DOI 10.1016/j.ymthe.2026.06.041, PMID 42427029) demonstrating that both A-A-V-CIB2 (semicircular canal delivery in Cib2 mutant mice at P0-P4) and A-A-V-CIB3 (intracochlear at P0-P1) rescue stereocilia architecture and hearing thresholds in the apical-to-middle cochlear turns of a mouse model of DFNB48 non-syndromic prelingual deafness — a two-independent-payload demonstration that CIB2 and its paralog CIB3, which bind TMC1/2 via partially distinct interfaces, are functionally interchangeable at the mechano-electrical-transduction channel apparatus, opening a "MET-complex module" commercial thesis in target-white-space (every commercial hearing-loss gene therapy program in the world today — Regeneron/Decibel, Eli Lilly/Akouos DB-OTO, Sensorion SENS-501 — targets OTOF, which is auditory neuropathy at the base of the hair cell; CIB2/CIB3 DFNB48 is a mechano-transduction defect at the top of the hair cell that no commercial vehicle currently addresses) that lands directly on top of Aletira Therapeutics' hereditary-hearing-loss + SELEXON cell-type-selective genetic-medicine thesis, giving Blackbird a payload-shopping opportunity inside the UM Ventures × Blackbird co-investment surface that produced aSKY; plus Pick 2 Ernst lab (UMB School of Dentistry Microbial Pathogenesis) BECC470s synthetic TLR4 adjuvant platform double-shot (Vaccine 2026-05-31 seasonal H1N1 + mSphere 2026-07-10 H5N1 pandemic-preparedness) demonstrating 18-month durable memory + T-helper-1-skewed IgG2a + complete homologous 2004 H5N1 protection + antibody breadth to HA head + stalk — a licensable T-L-R-4 adjuvant chassis carried by an NIH/NIAID adjuvant-development contract 2023-2028 competitive against Dynavax CpG-1018 + G-S-K MPL/AS01 + Novavax Matrix-M; plus Pick 3 Merchenthaler (UMB SoM Epi & Public Health) + Prokai (UNTHSC Pharmacology) J Neurosci 2026-07-08 marmoset primate proof-of-concept for the DHED brain-selective 17-beta-estradiol prodrug in aromatase-inhibitor-treated non-human primates (chronic oral DHED + letrozole → brain E2 elevation with no peripheral estrogen exposure + preserved hippocampal memory + reduced sleep fragmentation), an 11-year-old joint-UMB-UNTHSC platform finally arriving at the diligence bar a translational neurotherapeutics seed investor asks to see, in the aromatase-inhibitor-nonadherence unmet-need surface across ER-positive breast cancer + menopausal cognition + HRT-averse populations

Sat, 11 Jul 2026 12:00:00 +0000

Sourcing Radar for Saturday, July 11, 2026. Three UMB items, four days — the strongest single-week UM Ventures pipeline signal in months. Sunday tomorrow is a Portfolio Watch day; today is Radar only. **Pick 1 (LEAD) — Rehman/Ahmed AAV-CIB2 + AAV-CIB3 hearing gene therapy for DFNB48.** Rehman S (first), Giese APJ, Dragich AK, Riazuddin S, Frolenkov GI, Ahmed ZM (senior corresponding; zmahmed@som.umaryland.edu). "Rescue of stereocilia architecture and hearing function by AAV-CIB2 and AAV-CIB3 in a mouse model of recessive deafness DFNB48," *Mol Ther*, DOI 10.1016/j.ymthe.2026.06.041, PMID 42427029, published 2026-07-09. UMB SoM Otorhinolaryngology (cEAR Lab; Ahmed/Riazuddin/Hertzano) + Molecular Biology & Biochemistry + Ophthalmology & Visual Sciences. Frolenkov lab (Kentucky Physiology) as biophysics collaborator. **Working from abstract only** — paper is behind Elsevier paywall (Cell Press Molecular Therapy) with no bioRxiv preprint located; the abstract is substantive and the paper is the clearly dominant pick, so per PIPELINE.md we proceed with abstract-level summary and flag the depth. **Findings.** (1) AAV-CIB2 delivered via semicircular canal injection into Cib2-mutant mouse pups produces sustained, partial recovery of hearing thresholds + restoration of stereocilia architecture at the hair-bundle level in apical-to-middle cochlear turns; window is P0-P4 (earlier = better). (2) Intracochlear AAV-CIB3 at P0-P1 also restores hearing at apical-to-middle frequencies in the same Cib2-mutant model — module-level rescue rather than gene-specific, because CIB2 and CIB3 both bind TMC1/2 (pore-forming subunits of the MET channel) via partially distinct interfaces. Two independent payloads. **So what for Blackbird.** Direct Aletira portfolio adjacency. Aletira Therapeutics — Blackbird's flagship spinout, first fully-incubated program (nonprofit grant → seed → BioHub bench), Geoffrey Lynn CEO — is a hereditary-hearing-loss + cell-type-selective genetic medicine company. Its SELEXON platform uses alternative RNA splicing to restrict transgene expression to target cell types, and it's a bolt-on to AAV. The Ahmed paper delivers a payload (AAV-CIB2 or AAV-CIB3 for DFNB48); Aletira has a natural technology fit as a cell-type restriction chassis on top of that payload — restricting expression to inner + outer hair cells, avoiding vestibular sensory epithelium + supporting cells. Target-white-space: every commercial hearing-loss gene therapy program targets OTOF (Regeneron/Decibel; Eli Lilly/Akouos DB-OTO with positive Phase 1/2 data 2024; Sensorion SENS-501). CIB2/CIB3 DFNB48 is a mechano-transduction-apparatus defect at the top of the hair cell — no commercial vehicle currently addresses it. Regulatory + delivery route is clinically de-risked by the OTOF programs. UM Ventures × Blackbird co-investment surface (same track as aSKY). **Actions.** Esther — UM Ventures IP-status touchpoint on CIB2/CIB3 patent family (composition-of-matter + methods-of-use for gene therapy); flag two-payload optionality (CIB2 for its own disease + CIB3 as module-level rescue extending to CIB-family-adjacent hearing loss genotypes and possibly other MET-complex defects). Hemaka — courtesy touch to Zubair Ahmed via cEAR Lab; framing = "Aletira technology thesis has a natural pairing here." Eddie — file alongside aSKY under UM Ventures × Blackbird co-investment surface + Aletira portfolio adjacency; commercial diligence memo on the DFNB48 patient population size + OTOF gene-therapy market comparables (Akouos Lilly acquisition 2022, Decibel Regeneron 2023). Avi — chemistry not applicable; monitor Ahmed lab bench for follow-on preprints on capsid engineering + PROM1/OTOA/other MET-complex targets. **Watch.** Ahmed disposition on spinout vs. licensing; UM Ventures posture on the composition-of-matter estate; whether the cEAR Lab (Ahmed + Riazuddin + Hertzano) has other MET-complex targets in the pipeline that could aggregate into a company platform. **Pick 2 — Ernst lab BECC470s synthetic TLR4 adjuvant platform (two-paper double-shot).** Ernst RK (senior corresponding on both; rkernst@umaryland.edu). UMB School of Dentistry, Department of Microbial Pathogenesis. Paper A — Das S, Tenaglia BM, Riley D, Speed S, Baracco L, Gardner FM, Varisco DJ, Yang H, Nijhuis H, Kerstetter L, Krammer F, Sun W, Coughlan L, Frieman MB, Ernst RK (senior), "BECC-adjuvanted hemagglutinin influenza vaccine promotes enhanced immunogenicity and protective efficacy," *Vaccine* 87:128774, DOI 10.1016/j.vaccine.2026.128774, PMID 42218860, published 2026-05-31. Paper B — Riley DV, Baracco L, Das S, Tenaglia BM, Speed S, Dillen C, Hayes J, Del Veliz S, Nijhuis H, Le Sage V, Despres HW, Coughlan L, Sun W, Frieman MB, Ernst RK (senior), "Optimizing an avian influenza vaccine using a novel Bacterial Enzymatic Combinatorial Chemistry (BECC) TLR4 adjuvant," *mSphere* e0017126, DOI 10.1128/msphere.00171-26, PMID 42429648, published 2026-07-10. Krammer + Sun (Icahn/Mt. Sinai Microbiology) + Coughlan + Frieman (UMB SOM Center for Vaccine Development & Global Health) as collaborators. **Platform.** Bacterial Enzymatic Combinatorial Chemistry = engineered lipid-A analogs targeting TLR4 innate-immune receptor. BECC470s is the fully-synthetic lead-optimized molecule. Two BECC molecules (438 + 470s) are in the formal NIAID Adjuvant Development Program pipeline; Ernst has an active NIH/NIAID contract 2023-2028 for IND-enabling studies on BECC470s. **Findings A (seasonal H1N1).** BECC470s + recombinant HA in mice → strong T-helper-1 skew, robust IgG2a subclass responses, rapid viral clearance by day 7 post-challenge, outperforms MPL + PHAD comparators, and — headline metric — 18-month durable immune memory. **Findings B (H5N1 pandemic prep).** BECC470s + recombinant H5-HA in mice → complete survival vs homologous 2004 H5N1 challenge (prime-boost); antibody binding broadens to both HA head + stalk; neutralization remains strain-specific (does not extend to 2022 or 2024 H5N1 isolates). Real limitation the next optimization cycle needs to close; does not diminish platform story. **So what for Blackbird.** Licensable TLR4 adjuvant chassis inside UMB — specifically inside the School of Dentistry, a campus location a translational venture scan would not prioritize by default. Competitive against Dynavax CpG-1018 (Heplisav-B), GSK MPL + AS01 system, Novavax Matrix-M. Not a spinout candidate today; more likely a licensing carve-out or an asset drop into a vaccine-adjuvant-focused NewCo when the NIAID IND-enabling package matures. Federal non-dilutive money already carrying the tox + manufacturing work any commercial partner would want to see completed. **Actions.** Esther — UM Ventures IP-status probe on BECC470s composition-of-matter estate (Ernst is inventor); flag NIAID contract structure (typically retains university rights, but confirm). Eddie — landscape scan of adjuvant consolidator plays; Novavax, Emergent BioSolutions, GSK Vaccines, Dynavax as potential partners or acquirers. Hemaka — courtesy touch to Ernst; framing = "curious about the NewCo carve-out structure that would make sense post-IND-enabling." **Pick 3 — Merchenthaler/Prokai DHED brain-selective estradiol prodrug marmoset primate POC.** Cournoyer H, Duenas AM, Bharadwaj A, Kania N, Burns I, Regan K, Sweet J, Amato J, Saia C, Sabbouh G, Bergman J, Nguyen V, Remage-Healey L, Vazey EM, Merchenthaler I, Prokai L, Lacreuse A (senior corresponding). "Brain-Selective Estrogen Therapy in Male and Female Marmosets Partially Counteracts the Adverse Effects of Aromatase Inhibition on the Brain and Behavior," *J Neurosci* 46(27), DOI 10.1523/JNEUROSCI.2021-25.2026, PMID 42259621, published 2026-07-08. Corresponding author: Agnès Lacreuse at UMass Amherst Psychological & Brain Sciences (lacreuse@umass.edu). Merchenthaler at UMB SoM Epidemiology & Public Health; Prokai at University of North Texas Health Science Center Pharmacology & Neuroscience. **Molecule.** DHED = 10β,17β-dihydroxyestra-1,4-dien-3-one — brain-selective E2 prodrug. Enzymatically converted to 17β-estradiol only in brain tissue (aldo-keto reductase 1C family); inert in periphery (liver, uterus, breast). Original Sci Transl Med publication Prokai/Merchenthaler 2015 in rodent models. **This paper is the primate POC.** Chronic oral DHED in male and female marmosets treated with letrozole (aromatase inhibitor; models the estrogen-deprived state that adjuvant ER+ breast cancer patients experience) → (1) robust increase in brain 17β-estradiol across brain regions without affecting peripheral estrogen; (2) improved memory at short delays; prevented letrozole-induced cognitive slowing in a hippocampal-dependent memory task; (3) normalized hippocampal neuronal membrane potential + excitability; (4) reduced sleep fragmentation; (5) sex-differentiated thermoregulation effects (opposite direction in M vs F) — flagged for additional research. **So what for Blackbird.** First-in-class, class-of-one CNS-selective estrogen prodrug in a defined unmet-need patient population: AI-associated cognitive/sleep side effects drive nonadherence in ER+ breast cancer adjuvant hormonal therapy (large defined population; peak nonadherence is a well-documented clinical problem). Extensible to menopausal cognition + HRT-averse populations. Primate POC is the diligence bar for translational neurotherapeutics investors. 11-year-old platform (2015 Sci Transl Med) that has been publishing steadily but not moved into a company vehicle — this is the calendar moment for a spinout conversation. UMB-UNTHSC joint IP; Merchenthaler + Prokai named inventors. **Actions.** Esther — UM Ventures IP-status probe on the DHED estate; whether there's an active option agreement in place with the North Texas side. Hemaka — courtesy touch to Merchenthaler + Prokai via UMB SoM Epi & Public Health. Eddie — landscape scan of AI-associated cognitive-side-effect programs (largely empty; SERMs are peripheral); commercial-model diligence on the ER+ breast cancer adjuvant HRT-substitute market. **Watch.** Whether UMB tech-transfer is actively marketing the DHED asset or waiting for inbound interest; whether Prokai has parallel entrepreneurship footprint at UNTHSC that could serve as the founding operator. **Editorial framing.** UM Ventures pipeline delivery day. Three UMB translational papers in four days: School of Medicine hair-cell gene therapy (Aletira-adjacent), School of Dentistry TLR4 adjuvant chemistry (licensable platform), School of Medicine CNS endocrinology (first-in-class primate POC). All three sit inside the UM Ventures × Blackbird co-investment surface — same track that produced aSKY. Ahmed CIB2/CIB3 is the highest-value item — direct Aletira portfolio adjacency, genuine target-white-space in a clinically-de-risked modality. Ernst BECC470s is a platform-IP block with federal money already carrying IND-enabling work. Merchenthaler-Prokai DHED is a calendar-moment diligence-bar-cleared readout on a decade-old platform. Paper links: https://doi.org/10.1016/j.ymthe.2026.06.041 ; https://doi.org/10.1016/j.vaccine.2026.128774 ; https://doi.org/10.1128/msphere.00171-26 ; https://doi.org/10.1523/JNEUROSCI.2021-25.2026

Sourcing Radar — a rare three-way schizophrenia convergence day across two partner institutions: U-M-B Maryland Psychiatric Research Center medRxiv paper 2026-07-07 delivers a 58-patient double-blind placebo-controlled tryptophan-challenge crossover study giving human proof-of-mechanism for kynurenine aminotransferase II (KAT II) inhibition — 15g N-acetylcysteine pre-treatment significantly reduces tryptophan-induced serum KYNA rise (p=0.002), blunts white-matter cerebral blood flow elevations, and preserves the MATRICS cognitive composite score against placebo worsening (p=0.04) — Robert Schwarcz senior PI + Buchanan senior corresponding, and Schwarcz's COI declares him a scientific cofounder of Kynexis B.V. (Naarden, NL; €57M Series A led by Forbion Nov 2023) which licensed KYN-5356 from Mitsubishi Tanabe and is in Phase 2 for CIAS + AD expansion — a UM Ventures playbook lesson (the KAT II franchise Blackbird could have owned went offshore because UMB had the biology but no molecule to compete for the license) directly transferable to the cathepsin B CAR-T adjunct thesis under Luetkens; plus back-to-back Wednesday Lieber Institute press releases surfacing (a) Rossi/Sportelli/Weinberger/Pergola Nature Genetics paper 2026-06-22 introducing INGENE + MODULE trans-eQTL frameworks that recover 766 SCZ-associated genes including 641 novel (84% novel-discovery rate) with pathway enrichment on glutamate signaling + immune + brain development — direct priority-hit-sheet infrastructure for the Lieber GPR52 NewCo and a patentable target-generation substrate applicable across all PGC wave-3 disorders — and (b) Shah/Totty/Martinowich Cell Reports 45(6):117500 dorsal anterior cingulate cortex spatial-transcriptomics atlas confirming molecular signature of the L4-absent agranular architecture, nominating two novel L6 sublayers with divergent marker genes from dlPFC (KCTD8 for L6a; ADRA2A + CRHBP for L6b), and mapping von Economo neurons — schizophrenia + suicidal-psychosis + ASD-relevant spindle neurons — to deep L5 with mouse retroviral tracing inferring subcortical amygdala/VTA/pons connectivity, a second matched-donor Visium atlas platform substrate (post-dlPFC 2024) that strengthens the case for a Lieber-Institute × Kalhor 3DEEP deep-tissue spatial-omics cross-cycle collaboration

Fri, 10 Jul 2026 12:00:00 +0000

Sourcing Radar for Friday, July 10, 2026. Three schizophrenia papers, three days, two partner institutions (UMB + Lieber). Unusually clean convergence — all three change the read on Blackbird's Lieber neuropsychiatric NewCo (GPR52 program). **Pick 1 (LEAD) — UMB / Maryland Psychiatric Research Center: human proof-of-mechanism for KAT II inhibition in schizophrenia cognitive impairment.** Hare SM (first), Kelly DL, Pan Y, Chen S, Blatt F, Gorelick DA, Gold JM, Sathyasaikumar KV, Adhikari BM, Kochunov P, Wijtenburg SA, Rowland LM, Schwarcz R (P.I., NIMH 2P50MH103222 program-project grant), Buchanan RW (senior corresponding). "N-acetylcysteine and the Kynurenine Pathway: A double-blind, placebo-controlled, tryptophan challenge study in people with schizophrenia," medRxiv 2026.06.25.26356572, posted 2026-07-07. ClinicalTrials.gov NCT04013555. **Design.** 58 DSM-5 schizophrenia/schizoaffective patients, double-blind placebo-controlled crossover. Each patient given up to 15g NAC or placebo pre-treatment, then 30 min later 6g oral L-tryptophan (approved by FDA IND# 114931). Two challenge days ≥2 weeks apart. Serum kynurenine + KYNA, MATRICS Consensus Cognitive Battery (MCCB) composite, BPRS/SANS/CDS symptoms, and MRI (ASL, DTI, 1H-MRS) pre/post. **Rationale.** NAC directly inhibits KAT II in recombinant enzyme assay at IC50 ~500 μM (Sathyasaikumar/Schwarcz prior work). Oral tryptophan floods the kynurenine pathway on demand → forces peripheral + central KYNA production → NAC pre-treatment is the KAT II probe. **Results.** (1) NAC significantly reduced tryptophan-induced serum kynurenine (t=-2.02, p=0.048) and KYNA (t=-3.21, p=0.002) at 240 min. (2) Whole-brain white-matter CBF elevations were significantly blunted (t=-2.15, p=0.037); gray-matter CBF trend (p=0.078). (3) MCCB composite worsened on placebo, preserved on NAC (t=2.07, p=0.04); WMS-III Spatial Span working memory actually improved (t=2.41, p=0.02). (4) No effect on H-MRS glutamate/glutathione or symptom measures. (5) Safety clean (12 non-serious AEs, evenly distributed). **Genetic sub-analysis.** Trend for stronger NAC response in KMO rs2275163 CC-genotype patients (Cohen's d=0.51, p=0.095) — patient-selection hypothesis for a purpose-built KAT II inhibitor. **COI signal.** Robert Schwarcz disclosed as scientific cofounder of **Kynexis B.V.** (Naarden, Netherlands; launched Nov 2023 with €57M Series A led by Forbion + Ysios Capital + Sunstone). Kynexis has KYN-5356 (first-in-class KAT II inhibitor, licensed from **Mitsubishi Tanabe Pharma**) in Phase 2 for CIAS + expanding into AD. Buchanan advisory-board Merck/Newron/Roche + Karuna. Kelly advisory-board Teva/BMS/BI + Alkermes consultant. **So what for Blackbird.** Three layered reads. (1) **Validation.** Human PoM for a non-dopaminergic, non-muscarinic path to CIAS — a tailwind for the Lieber GPR52 program on same disease + same symptom domain (all-three-symptom-domains narrative). (2) **Competitive clock.** KYN-5356 Phase 2 read-out will land ahead of the Lieber NewCo DC declaration expected later 2026. If it works, KYN-5356 becomes first-in-class in the space and GPR52 differentiation story has to sharpen. (3) **UM Ventures playbook lesson.** UMB has world's leading KAT II biology (Schwarcz) + world's leading KAT II trialist (Buchanan) under one roof + NIH P50 program-project funding + fresh human PoM data — but the KAT II franchise is at a Netherlands VC-led company because Mitsubishi Tanabe had the molecule and UMB tech-transfer did not. **Directly transferable warning to the cathepsin B CAR-T program under Luetkens** — same UM Ventures × Blackbird co-investment surface that produced aSKY, similar "great biology, need the molecule" risk shape. **Actions.** Esther — UM Ventures touchpoint on cathepsin B COM/methods-of-use IP posture (this week; framing: "we saw what happened on KAT II"). Hemaka — Luetkens conversation about mediation of a small-molecule discovery arm collaborating with Fannie Angelos GMP Lab. Eddie — track KYN-5356 Phase 2 readouts against GPR52 DC calendar; Kynexis is the primary comparator for the Lieber NewCo pitch narrative. Avi — brief scan of the KAT II selective-inhibitor chemistry space (competitive intelligence + adjacent-target scoping). **Pick 2 — Lieber Institute / Third Rock coexpression-based schizophrenia TWAS platform (Nature Genetics).** Rossi F, Sportelli L, Kikidis GC (co-first, Univ Bari + Lieber), ..., Del-Ben CM, Thibaut F, Weinberger DR (co-senior), Pergola G (senior corresponding). "Co-expression-based models improve eQTL predictions for transcriptome-wide association studies and highlight new schizophrenia-associated genes," *Nature Genetics*, DOI 10.1038/s41588-026-02646-3, published 2026-06-22 (surfaced via LIBD press release 2026-07-08). **Method.** INGENE and MODULE — two trans-eQTL frameworks sitting on top of standard cis-only TWAS. **INGENE.** Uses one gene's genetically-predicted expression to predict distant coexpressed partners — captures cis-to-trans regulatory hops through coexpression networks. **MODULE.** Identifies SNPs associated with the first principal component (eigengene) of gene coexpression modules — captures coordinated module-level regulation. **Data.** RNA-seq from six post-mortem brain regions (amygdala, caudate nucleus, dorsal + subgenual anterior cingulate cortex, dlPFC, hippocampus). Model integration significantly improves gene-level expression imputation (MLE α=0.05) for 18,744 genes across regions vs cis-only or EpiXcan benchmark. **Application to schizophrenia GWAS.** PGC wave-3 SCZ genotypes → 766 SCZ-associated genes (693 non-MHC) at pFDR<0.01. **641 of those (83.7% novel-discovery rate) are new to schizophrenia TWAS literature.** 51% of novel associations come from trans-only predictions — long-range regulation is a large fraction of the SCZ genetic architecture that cis-only TWAS could not capture. Pathway enrichment: glutamate signaling, brain-cell communication, immune processes, brain development. **So what for Blackbird.** Priority-hit-sheet infrastructure for the Lieber GPR52 NewCo + second-and-third-asset planning. Two commercial angles. (1) **The 641-novel-gene target list** is a licensable substrate for the NewCo's pipeline expansion beyond GPR52 — glutamate-signaling + brain-development + immune buckets are the natural adjacencies. (2) **The INGENE + MODULE frameworks themselves are patentable computational methods** applicable across all PGC wave-3 disorders (bipolar, MDD, ASD, OCD) — that's a target-generation engine that could seed multiple NewCo pipeline assets. **Actions.** Hemaka — courtesy touch to Weinberger + Pergola via Third Rock relationships. Framing: "Blackbird has scaling interest in the INGENE-MODULE substrate as a target-generation asset for the neuropsych NewCo, separate from the SCZ hit list." Esther — JHTV probe on whether the INGENE + MODULE frameworks are separately licensable, or bundled with the NewCo IP stack. Avi — file the 641-gene list against the NewCo's existing hit-to-lead pipeline; flag hits that intersect with GPR52-adjacent mechanisms as second-asset candidates. **Pick 3 — Lieber Institute matched-donor dACC spatial-transcriptomics atlas (Cell Reports).** Shah K, Totty MS (co-first; JHU Biostatistics), Bach SV, Valentine MR, Chandra A, ..., Kwon SH, ..., Page SC, Maynard KR, Hicks SC (co-senior; JHU Biostatistics + Malone Center for Engineering in Healthcare), Martinowich K (senior corresponding). "Spatio-molecular gene expression reflects dorsal anterior cingulate cortex structure and function in the human brain," *Cell Reports* 45(6):117500, DOI 10.1016/j.celrep.2026.117500, published 2026-06-04 (surfaced via LIBD press release 2026-07-08). **Design.** Paired snRNA-seq (Chromium; n=10 libraries) + Visium spatial transcriptomics (n=17 capture areas) on dorsal anterior cingulate cortex from 10 adult neurotypical donors — **same 10 donors** previously profiled for the Lieber dlPFC atlas (Huuki-Myers 2024 Science). Non-negative matrix factorization (NMF, k=75) to project cell-type-specific gene expression programs from snRNA-seq into SRT. **Three findings.** (1) **L4 absence confirmed at spatial-molecular resolution** — canonical L4 markers RORB, PVALB collapse into L5 in the agranular dACC (contrast with dlPFC in same donors). (2) **Two novel L6 sublayers.** L6a marker KCTD8, L6b markers ADRA2A + CRHBP — divergent from dlPFC L6, new territory for cognitive-control target discovery. (3) **Von Economo neurons mapped to deep L5.** NMF61 pattern is dACC-L5-ET-specific (VAT1L, SULF2, HAPLN4, FEZF2, GABRQ, POU3F1 markers). smFISH confirmation. Mouse retroviral tracing infers VEN axonal projections to amygdala, VTA, pons + prefrontal cortex — first molecular-level connectivity read on the VEN population. VEN loss is documented in schizophrenia + suicidal psychosis + ASD. **So what for Blackbird.** Platform-substrate, not direct spinout. Two commercial angles. (1) **Second matched-donor Visium atlas from Lieber in a year** (post-dlPFC 2024) — strengthens the case for a **Lieber × Kalhor 3DEEP** deep-tissue spatial-omics cross-cycle collaboration; 400-μm-depth 3DEEP + Lieber donor collection is the natural next-cycle platform paper. (2) **dACC atlas nominates the tissue block for target discovery in cognitive control + reward-processing + pain** — the three symptom domains where the Lieber NewCo pipeline should be probing for second-and-third assets. Open data + interactive portal at research.libd.org/spatialdACC/. **Actions.** Hemaka — introduce Kalhor (JHU BME 3DEEP) to Martinowich + Hicks; the deep-tissue dACC atlas is a Cell paper waiting to be written. Esther — cross-reference this atlas with any upcoming NewCo target-ID work planned by the neuropsych discovery team. **Editorial framing.** Rare three-way schizophrenia convergence day. Three papers, two partner institutions, all changing the read on Blackbird's Lieber neuropsych NewCo. UMB validates the mechanism space (non-dopaminergic paths to CIAS work in humans) but the KAT II franchise is with Kynexis — playbook lesson for cathepsin B under Luetkens. Weinberger's Nature Genetics paper delivers the target-generation engine + 641-gene priority-hit list. Martinowich's Cell Reports atlas delivers the anatomical substrate for cognitive-control target discovery. Paper links: https://www.medrxiv.org/content/10.64898/2026.06.25.26356572v1 ; https://www.nature.com/articles/s41588-026-02646-3 ; https://www.cell.com/cell-reports/fulltext/S2211-1247(26)00578-4

Sourcing Radar — a lean Thursday after Monday-through-Wednesday cleared the strongest translational items — two Hopkins peer-reviewed papers with platform-level significance for Blackbird's CNS thinking (Huganir's Snyder Neuroscience PNAS paper introducing sex- and experience-dependent turnover of long-lived synaptic proteins including GluA4 as the slowest AMPA subunit, ECM proteins as the memory-scaffolding cast, sex-differentiated Gabrg2 stability, and fear-conditioning-driven Shank3 stabilization — a novel target-selection lens for CNS drug discovery arguing sex has to be a primary axis of pharmacodynamic biomarker design for the GPR52 Lieber NewCo and any Snyder-adjacent psychiatric program; and Jiou Wang's Packard-funded JHU BSPH+Neuroscience PLoS Genetics multi-region transcriptomic atlas across 239 ALS patients establishing a subtype-and-sex-specific stratification framework — C9orf72 vs non-C9 diverge on lipid/immune vs protein-synthesis pathways, female non-C9 patients have significantly higher cervical-cord microglial burden that correlates with shorter clinical duration — the patient-stratification layer that every TDP-43-directed program including the Wong lab caspase-3-tau axis will need to build into its clinical development plan, third data point in a Packard-anchored TDP-43 arc this desk has walked through the past week) plus a signal item on Lieber Institute's Carr/Barrow behavioral pharmacology core publishing post-weaning social isolation adolescent reward-seeking phenotypes in Physiology & Behavior — a read on where Blackbird's own GPR52-NewCo platform partner's scientific attention sits right now (environmental stress + adolescent development + reward biology, the frontier a next-generation schizophrenia or bipolar program has to speak to)

Thu, 09 Jul 2026 12:00:00 +0000

Sourcing Radar for Thursday, July 9, 2026. Honest lean-day framing: after Monday-through-Wednesday cleared out the strongest translational items from the past 10 days (07-06 AsclepiX suprachoroidal minipig tox + OptaNova disclosure; 07-07 Wong lab caspase-3-tau axis; 07-08 triple pick of Watkins ACSVL3/grassofermata GBM + Cahan-Raman DDX3/RK-33 pediatric OS + Luetkens cathepsin B CAR-T trogocytosis), the Baltimore preprint funnel has quieted. Today's three items are platform-signal reads, not licensable sourcing leads — "what are the marquee labs working on" rather than "would a translational investor take this meeting." **Pick 1 (LEAD, platform signal) — Huganir lab Snyder Neuroscience PNAS paper.** S.H., D.-G.M., A.P., A.M.B., R.L.H., "Sex- and experience-dependent regulation of synaptic protein turnover," *Proc Natl Acad Sci USA*, published 2026-07-07, DOI 10.1073/pnas.2602111123. Solomon H. Snyder Department of Neuroscience, Johns Hopkins University School of Medicine. **COI: none.** Funding: NIH R01MH112152 + R37NS036715 (both to RLH); R00MH124920 (AMB); U01CA271410 + P30CA15083 (AP). **Method.** Stable-isotope labeling in mammals (SILAM) — pulse Lys-13C6 heavy diet through breeding pair + offspring to 10 wk, chase Lys-12C6 light diet, LC-MS on hippocampal PSD fractions across timepoints, exponential-decay fit to relative isotope abundance yielding protein half-lives across 4,145 PSD proteins. **Long-lived synaptic protein (LLP) set.** Extracellular matrix components dominate the top of the stability distribution — Vcan (versican), neurocan, hyaluronan-and-proteoglycan link proteins 1+4, Traf3, Rras2, CRMP family members. Huganir's own AMPA-receptor subunit GluA4 is identified as the slowest-turnover AMPA subunit; Sap102 (MAGUK family) is the fastest. Maps neatly onto the perineuronal-net memory-scaffolding literature — mature-brain synapses build physical structures meant to persist and provide long-term-plasticity substrate. **Sex-differentiated finding.** Gabrg2 (GABA-A γ2 subunit; ASD risk gene) shows significantly greater stability in male mice vs female mice. Multiple additional ASD-risk genes (Cdkl5, Taok1, Shank3) show sex-dependent stability. Threshold for classification: male/female half-life ratio >2 SD from mean. **Experience-dependent finding.** Contextual fear conditioning (single session, 5 × 0.75 mA 2-s foot shocks) stabilizes Shank3, another ASD/neurodevelopmental risk gene — presumably part of the memory-consolidation machinery. GO enrichment post-cFC: elevated stability of learning + synaptic organization + dendritic spine + PDZ-domain-binding proteins; decreased stability of "positive regulation of protein phosphorylation." Authors are candid they cannot fully disambiguate learning-dependent vs stress-induced changes. **Why it matters for Blackbird — psychiatric-drug-discovery lens.** Not a paper that spins out a company; Huganir has no COI declared and this is pure federal-grant-funded mechanism. But it changes two things about how a translational neuroscience investor should think. First, epidemiological sex differences in psychiatric risk (ASD more common in males, MDD more common in females) may be traceable to sex-differentiated stability of specific PSD proteins — testable molecular hypothesis for observations that have been epidemiological until now. Second, any schizophrenia or ASD or intellectual-disability program needs sex-stratified dosing-biomarker development + trial stratification as a primary axis. Directly relevant to Blackbird's Lieber neuropsych NewCo (GPR52 schizophrenia program) — PD biomarkers coupled to synaptic function will read differently in male + female brains, and the differences may be substrate-driven not just receptor-density-driven. Snyder Neuroscience is the marquee JHU neuroscience group — worth a coffee with Huganir. **Actions.** Esther — light JHTV probe on Snyder Neuroscience recent IP; whether the Huganir lab has any consulting/advisory-board footprint worth mapping. Hemaka — courtesy touch to Huganir via Snyder Neuroscience network; pitch = "the sex-differentiated LLP finding matters for the way Blackbird thinks about the Lieber NewCo dosing/biomarker strategy, would welcome a coffee." This is platform-visibility work, not licensing work. Not this quarter's deal — this quarter's relationship. **Pick 2 (arc-extender, platform framework) — Wang lab JHU BSPH+Neuroscience PLoS Genetics ALS transcriptomics.** Jiou Wang (senior corresponding), et al., "Multi-regional transcriptomic profiling reveals divergent molecular mechanisms in ALS-related neurodegeneration," *PLoS Genet*, published 2026-07-08, DOI 10.1371/journal.pgen.1012225. JHU Bloomberg School of Public Health, Dept of Biochemistry and Molecular Biology + JHU School of Medicine, Dept of Neuroscience. **COI: none. Packard Center for ALS Research at Johns Hopkins funded.** **Design.** 316 subjects total (77 controls, 239 ALS patients: 43 C9orf72+ "ALS-C9," 196 without mutation "ALS-non-C9"), transcriptomic profiling across 10 CNS regions — cerebellum, cervical/thoracic/lumbar spinal cord, lateral + medial motor cortex, frontal + hippocampus + occipital + temporal cortex. Integrative pathway + cell-type-deconvolution analysis. **Two findings.** (1) **Molecular subtype divergence.** ALS-C9 enriches for lipid-homeostasis + immune-modulation pathways; ALS-non-C9 enriches for protein-synthesis pathways. Divergence has been suspected but not cleanly mapped across regions. Short-duration ALS-C9 cases show upregulated immune + metabolic pathways; short-duration ALS-non-C9 cases show the opposite pattern — downregulation. (2) **Sex-within-subtype.** Female ALS-non-C9 patients have significantly higher microglial cell-type deconvolution scores than male counterparts in the cervical spinal cord — the burden correlates with shorter clinical duration. Female non-C9 ALS may be a distinct pharmacological population from male non-C9 ALS. **Sourcing significance — arc completion.** The Packard Center is the exact Hopkins hub Blackbird already had visibility on via the 07-04 Packard imaging item. This paper is the third data point in a week for a Packard-anchored TDP-43-adjacent thesis: 07-04 (Packard imaging platform), 07-07 (Wong lab caspase-3-tau mechanism reframing addressable market from ALS+FTD into AD/ADRD), 07-08 mention (context for Wong lab), and now this Wang lab framework paper. Argument: TDP-43-directed therapeutic development needs subtype-and-sex-stratified frameworks. That's the patient-stratification layer any TDP-43-directed program (Ionis-Amylyx STMN2 ASO, QurAlis, Trace Neuroscience, or a Wong-lab-derived caspase-3-tau asset) will need to build into its clinical development plan. **Actions.** Esther — file paper alongside Wong lab pipeline as a landscape framework for TDP-43 platform stratification biology. Hemaka — check whether Jiou Wang would be open to a translational grant conversation with Blackbird on the next natural question (e.g., cell-type-resolved MRPP-scale profiling in TDP-43-co-pathology AD; if yes, this is a Blackbird-shaped translational-grant target). Eddie — cross-referenced against Wong lab + Packard imaging as three data points on a coherent Hopkins TDP-43 platform thesis. Reading list for Esther+Hemaka on Blackbird's Packard visibility. **Pick 3 (signal, not lead) — Lieber Institute Carr+Barrow behavioral pharmacology.** Noback M, Husaini ZA, van Kralingen T, Afzal A, Li Y, Barrow JC, Carr GV (senior corresponding; greg.carr@libd.org), "Post-weaning social isolation increases reward-seeking behavior in mice," *Physiol Behav*, published 2026-07-04, DOI 10.1016/j.physbeh.2026.115437. Lieber Institute for Brain Development + JHU Department of Physiology, Pharmacology & Therapeutics. **What.** Post-weaning social isolation (PND 21-35) in mice increases performance on touchscreen rCPT + higher breakpoints on PR4 + more reinforcers earned on FR1 in adulthood — the improvement is motivational, not attentional. Straight behavioral neuroscience — no target, no drug, no mechanism nomination. **Why it's worth 90 seconds of the episode.** James Barrow (LIBD translational medicinal chemistry) + Gregory Carr (LIBD behavioral pharmacology core) are the medchem + behavioral duo running Blackbird's GPR52 schizophrenia NewCo platform. That they're authoring a paper on adolescent-stress-driven motivational phenotypes tells you where the Lieber platform's scientific attention is right now: the intersection of environmental stress, adolescent development, and reward biology — three domains that increasingly define the frontier a next-generation schizophrenia or bipolar program has to speak to. **Read as a snapshot of your own platform partner's core.** Not a sourcing lead; not a company-formation conversation. A read on the shop. **Editorial framing.** Two Hopkins peer-reviewed papers + one Lieber signal item. Nothing spins out this week. What shapes the next few weeks: sex-stratified thinking in CNS drug development (Huganir), TDP-43 subtype-and-sex patient-stratification biology (Wang + Packard-anchored arc), and Lieber's core interests as the platform partner (Carr+Barrow). Thin days are healthy pipeline; not every day has a licensable asset. Paper links: https://www.pnas.org/doi/10.1073/pnas.2602111123 ; https://doi.org/10.1371/journal.pgen.1012225 ; https://doi.org/10.1016/j.physbeh.2026.115437

Sourcing Radar — Watkins lab (Kennedy Krieger + JHU Neurology) bioRxiv 2026-07-08 preprint nominates a first-in-class metabolic target for glioblastoma: very long-chain acyl-CoA synthetase 3 (ACSVL3) is selectively expressed in glioma vs normal glia, genetic depletion collapses malignancy in U87MG + Mayo-22, natural-product tool compound grassofermata/CB5 reproduces the phenotype (differentiation via GFAP induction + reduced FA β-oxidation + attenuated invasion) with 3 μM efficacy + 10 μM tumor-selective toxicity, 2 mg/kg/day IP shows tumor regression in U87MG NOD/SCID xenografts at >16x therapeutic window (32 mg/kg/day tolerated); plus a Cahan/Raman JHU BME + SKCCC bioRxiv preprint layering JHU-owned DDX3 inhibitor RK-33 with mifamurtide in immunocompetent osteosarcoma models delivering MYC-status-selective metastatic control (non-MYC-amp responds; MYC-amp fails; biomarker-selectable pediatric bone-sarcoma small-molecule immuno-combo); plus a UMB Luetkens lab Signal Transduction and Targeted Therapy paper identifying cathepsin B as required for CAR-mediated trogocytosis in CAR-T cells + demonstrating that cathepsin B inhibition preserves CAR-T persistence — an adjunct small-molecule thesis compatible with every commercial CAR-T without touching cell engineering, and directly adjacent to UMGCCC's in-flight CD229 CAR-T trial + the aSKY-adjacent UM Ventures × Blackbird co-investment surface

Wed, 08 Jul 2026 12:00:00 +0000

Sourcing Radar for Wednesday, July 8, 2026. Three picks — two Johns Hopkins small-molecule oncology preprints + one UMB CAR-T adjunct target published in a Nature-portfolio journal. All three deliver novel mechanism + in vivo POC + defensible partner-institution IP; all three read as investor-take-the-meeting shape. **Pick 1 (LEAD) — first-in-class metabolic target for glioblastoma (Kennedy Krieger + JHU Neurology).** Clay EM, Shi X, Kolar EA, Liu Y, Lal B, Watkins PA (senior, corresponding; watkins@kennedykrieger.org). "Toward pharmacologic therapy for glioblastoma: Characterization of the very long-chain acyl-CoA synthetase 3 (ACSVL3) inhibitor Grassofermata," bioRxiv v1 2026.07.07.736493, posted 2026-07-08. CC-BY-ND 4.0. **Competing interests declared: none.** Funding: NIH R01NS062043 (Watkins). **Novel target.** ACSVL3 is a fatty-acid-activating enzyme selectively expressed in gliomas (U87MG + Mayo-22 + additional glioma cell models) vs normal glial cells — a clean expression differential + tumor-selective vulnerability. Genetic knockdown or knockout collapses malignant phenotype: slowed growth, reduced invasion, increased GFAP expression (astrocyte-lineage differentiation marker). **Tool compound.** CB5 / grassofermata — natural-product-derived small-molecule ACSVL3 inhibitor. 3 μM = growth arrest + differentiation + reduced β-oxidation of C16-C24 fatty acids, reversible on washout; 10 μM = tumor-selective toxicity (glioma cells killed, normal human fibroblasts unaffected — clean therapeutic-window read this early). **Xenograft POC.** NOD/SCID subcutaneous U87MG xenografts, 2 mg/kg/day IP → slowed growth by day 7, tumor shrinkage by day 12. Tolerability up to 32 mg/kg/day with no visible adverse effects → >16x therapeutic window on efficacy dose. **Two candid caveats.** (a) Mayo-22 patient-derived xenograft (more clinically representative model) did NOT show statistically significant efficacy — meaning heterogeneity/patient-selection biology is real + first diligence question. (b) Subcutaneous model, not orthotopic — no direct BBB data; grassofermata is small enough that prior on brain penetration is reasonable but is a wet-lab question. **So what for Blackbird.** GBM has had no drug of consequence in 20 years — the field is a graveyard, exactly the shape of problem where a first-in-class metabolic vulnerability + a tool compound + a >10-year lab program on ACS-family biology + a Kennedy Krieger + JHU IP stack + an active NIH R01 should get a meeting. Differentiation-inducing MOA is a narrative thesis (astrocytic lineage commitment for GBM cells is investor-legible). Slots alongside aSKY oncology thesis as a metabolic-target platform play. **Actions.** Esther — JHTV touchpoint on ACSVL3 patent-family status (any provisionals filed on grassofermata analogues, ACSVL3-target composition-of-matter, or method-of-treatment claims). Hemaka — courtesy touch to Watkins via Kennedy Krieger + JHU Neurology network; pitch = "first-in-class GBM metabolic target with tool compound + xenograft data is exactly the shape Blackbird incubates." Avi — molecular-discovery scoping on grassofermata-analog chemistry space + ACSVL3 orthosteric-vs-allosteric binding characterization; is this a natural-product-tractable series or does hit-to-lead need de novo screening. Eddie — file alongside Slusher gut-restricted GCPII IBD as a JHU-anchored small-molecule oncology sourcing lead; probable Blackbird incubation path if Watkins is open to spinout structure. **Pick 2 — DDX3 helicase inhibitor + immunostimulant combo for pediatric bone sarcoma (JHU BME + SKCCC + Einstein/Montefiore).** Weil BR (first co-corresponding; Einstein), Peng H, ..., Cahan PA (JHU BME + INBT + SKCCC), Raman V (JHU Oncology + Radiology + Pharmacology; RK-33 inventor), Loeb DM (senior corresponding; Einstein/Montefiore). "Innate immune microenvironmental remodeling via DDX3 inhibition improves outcomes in pediatric bone sarcoma models," bioRxiv v1 2026.07.01.735844, posted 2026-07-01. **Mechanism dichotomy.** DDX3 is a helicase with opposite roles in tumor cells vs macrophages. In tumor cells DDX3 unwinds dsRNA and suppresses Type I IFN signaling; in macrophages DDX3 supports inflammatory cytokine expression. Inhibiting DDX3 flips both dials in the same therapeutic direction — dsRNA accumulation in tumor cells triggers Type I IFN response + broader inflammatory gene expression; tumor-associated macrophages polarize toward M1-like pro-inflammatory phenotype. **Combo POC + biomarker.** RK-33 (Hopkins-invented DDX3 small-molecule inhibitor from the Raman lab) + mifamurtide (Muromonab-adjacent immunostimulant, only approved OS immunotherapy — in Europe, not US) in immunocompetent osteosarcoma mouse models. **In non-MYC-overexpressing backgrounds, the combination significantly reduces metastatic burden. In MYC-amplified backgrounds, the combination fails.** MYC-status-based patient selection biomarker landing in the same package as the pharmacology. **So what for Blackbird.** Pediatric OS survival has been ~25% in metastatic disease for decades; there is a Hopkins-invented small molecule with prior clinical development history (RK-33 has been through Phase 1 in adult solid tumors) + a rational immuno-combo positioning + a genomic patient-selection biomarker. IP position: RK-33 composition-of-matter is inside the JHU estate. The immuno-combo indication + MYC-biomarker strategy is fresh. Pediatric oncology + orphan pricing + high unmet need = clean spinout narrative. **Actions.** Esther — JHTV touchpoint on RK-33 IP status (composition-of-matter, method-of-use for OS, MYC-status stratification claims). Hemaka — courtesy touch to Raman/Cahan via BME + SKCCC network; check whether RK-33 is currently licensed or unencumbered, and whether Raman is open to a company-formation conversation on the OS-immunocombo indication. Eddie — cross-check against active pediatric-oncology strategics (Y-mAbs, Aadi, Advaxis, Day One Biopharmaceuticals) as future partnering/M&A landscape. Slots alongside aSKY as a JHU-anchored oncology sourcing lead. **Pick 3 — cathepsin B as a druggable failure-mode target for CAR-T therapy (UMB Luetkens lab + UMD College Park Upadhyaya lab).** Dietze K (first), Nguyen K, Kiely EA, Upadhyaya A (UMD College Park), Owonikoko TK (UMGCCC Executive Director), Luetkens T (senior corresponding; UMSOM Microbiology + Immunology + UMGCCC Fannie Angelos Cellular Therapeutics GMP Lab Director R&D). "CAR-mediated trogocytosis limits CAR-T persistence and is blocked by cathepsin B inhibition," *Signal Transduction and Targeted Therapy* (Nature portfolio) s41392-026-02654-z, June 2026. Funding: Maryland Cigarette Restitution Fund CH-649-CRF, NCI P30CA134274 + R21CA297125, NIH R35 GM145313, NIAID T32 AI095190, American Cancer Society DBG-25-1434942-01-ET. **Mechanism.** Lattice light-sheet microscopy imaging shows CAR-T cells physically strip target-antigen fragments off tumor cell surfaces and deposit those fragments onto themselves — a form of CAR-mediated trogocytosis (CMT). Consequence is dual damage: (a) tumor cell loses target antigen and evades further killing (relapse mechanism), (b) CAR-T cell becomes shorter-lived (persistence mechanism). Both are on the leading list of hypotheses for why most CAR-T relapses happen within 5 years. **Druggable node.** Cathepsin B — lysosomal protease — is required for the trogocytosis mechanism. Blocking cathepsin B prevents CAR-T cells from stripping tumor antigen, preserves CAR-T persistence, and augments tumor control in animal models. **Not a cell-engineering fix — a small-molecule adjunct.** Cathepsin B inhibitors are a well-trodden medicinal-chemistry space (broad-spectrum cysteine-protease inhibitors are known; selective + cell-permeable + biologically stable series is the medchem question). Repurposing angle + de novo composition-of-matter angle both viable. **So what for Blackbird.** Layers onto every commercial CAR-T product without touching cell engineering — CD19 (Kymriah, Yescarta, Breyanzi, Tecartus), BCMA (Abecma, Carvykti), and the next-generation CD229 CAR-T that UMGCCC is running in its own first-in-human trial for B-cell lymphoma. UMB IP inside the UM Ventures pipeline — same co-investment agreement that produced aSKY. Adjacency for Krupnick's aSKY oncology team + explicit UMGCCC operational surface. **Actions.** Esther — UM Ventures touchpoint on cathepsin B IP status (methods-of-use for CAR-T combination + selective inhibitor composition-of-matter). Hemaka — courtesy touch to Luetkens via UMB Microbiology + Immunology + Fannie Angelos GMP Lab network; pitch = "CAR-T adjunct thesis is investor-legible and Blackbird BioHub incubation model fits." Avi — molecular-discovery scoping on selective cathepsin B chemistry space + tumor-tissue-permeable formulations. Eddie — file alongside aSKY in UM Ventures × Blackbird co-investment portfolio surface; potential adjunct-with-every-approved-CAR-T commercial framing. **Editorial framing.** Three institutions on the map today — Kennedy Krieger + Hopkins Neurology, Hopkins BME + SKCCC, UMB SOM + UMGCCC. Two Hopkins picks are small-molecule oncology with mechanism-first stories; UMB pick is a cell-therapy adjunct on the aSKY-adjacent co-investment surface. All three fit the shape Blackbird incubates. **The ACSVL3 lead is the strongest sourcing signal of the week** — first-in-class metabolic vulnerability + tool compound + xenograft POC + fresh preprint + Watkins lab is an experienced translational program with active NIH R01 funding + Kennedy Krieger institutional relationship for Blackbird. Paper links: https://www.biorxiv.org/content/10.64898/2026.07.07.736493v1 ; https://www.biorxiv.org/content/10.64898/2026.07.01.735844v1 ; https://www.nature.com/articles/s41392-026-02654-z

Sourcing Radar — Philip C. Wong lab (JHU Pathology + Neuroscience) Molecular Neurodegeneration paper places TDP-43 loss-of-function upstream of caspase-3 endoproteolysis of tau in mouse + human iPSC-neuron models, extending the TDP-43 platform from ALS/FTD into the AD/ADRD indication space — Wong holds foundational patent WO2016205615A1, current paper declares no competing interests + Wong is speaking at 2026 ALS Drug Development Summit, positioning a fresh JHTV provisional filing on the caspase-3-tau axis in TDP-43-deficient neurons as the JHTV-timing signal Esther + Hemaka should chase this week before national industry attention compresses the window

Tue, 07 Jul 2026 12:00:00 +0000

Sourcing Radar for Tuesday, July 7, 2026. One high-conviction JHU mechanistic paper. Baghel MS + Burns GD (co-first, contributed equally), Tsapatsis M, Peethambaran Mallika A, Cruz ALF, Cao T, Chen XK, De La Rosa I, Marx SR, Ye Y, Sun S, Li T (now NIA), Ling JP, Wong PC (senior, corresponding; **JHU School of Medicine Pathology + Neuroscience**; wong@jhmi.edu), "TDP-43 dysfunction facilitates the pathological conversion of tau," *Molecular Neurodegeneration* (BMC / Springer Nature), received 2026-04-16, accepted 2026-06-25, published 2026-07-04 (unedited early-access version), DOI 10.1186/s13024-026-00968-8 (PMID 42401929). Open access CC-BY. **Competing interests declared: none.** Funding: NIH R01NS095969, R33NS115161, R61NS115161 (all to PCW); Alzheimer's Drug Discovery Foundation GC-202402-2026079 (to PCW). Full text via r.jina.ai reader proxy on the BMC HTML (link.springer.com redirected to auth wall; molecularneurodegeneration.biomedcentral.com rendered via jina). **Prior JHU-adjacent TDP-43 dedup check (per PIPELINE re-invocation rule + weekly recall).** The 2026-07-04 episode covered **Jiou Wang** (JHU BMB + Neuroscience; Packard Center-funded) eLife noncanonical-amino-acid Anap TDP-43 live-cell-imaging platform + U-Arizona **Xinglong Wang / XL20** Nature Aging small-molecule TDP-43 conserved-region compound (JHU **Pan P. Li** lab as supporting coauthor). Today's Baghel/**Philip C. Wong** paper is a **different lab** (Philip Wong ≠ Jiou Wang, ≠ Pan P. Li), a **different scientific question** (tau proteotoxic downstream consequence of TDP-43 LOF, not TDP-43 imaging or TDP-43 chemistry), and a **different indication frame** (AD/ADRD rather than ALS). Legitimately fresh. **What the paper shows — mechanism.** Model 1 — genetic mouse. *Tau4R;CaMKII-CreER;Tardbp^f/f^* triple-genotype forebrain conditional model. Floxed *Tardbp* + tamoxifen-induced CaMKII-driven CreER selectively deletes TDP-43 in forebrain excitatory neurons; *Tau4R* transgene provides genetic tau-seeding substrate (four-repeat MBD of human tau). Model 2 — human iPSC cortical neurons. TDP-43 LOF induced in human iPSC-derived cortical neurons — TDP-43-dependent cryptic splicing readouts followed by caspase-3-mediated tau endoproteolysis measurements. **Temporal order: cryptic splicing precedes caspase-3-dependent tau cleavage.** Causal direction: TDP-43 LOF first → cryptic splicing → caspase-3 activation → tau proteolytic cleavage. **Mechanism claim in three steps.** (1) TDP-43 LOF in forebrain neurons exacerbates tauopathy-dependent brain atrophy in mice. (2) Vulnerable neurons specifically sensitive to caspase-3-dependent cleavage of endogenous tau; that cleavage generates a tau species that seeds further tauopathy. (3) Using the *Tau4R* genetic seeding approach, tau seed dose correlates with caspase-3-dependent tau cleavage, accelerated tauopathy, and loss of vulnerable neurons deficient in TDP-43. **Vulnerable-neuron concept.** Wong's lab has been building this across 2020-2026 literature — specific populations of forebrain excitatory neurons in entorhinal cortex, hippocampus, cortex (the exact populations that die in AD) are selectively sensitive. This paper extends into a two-hit interaction: tau seed × TDP-43 LOF = accelerated proteotoxicity + selective vulnerable-neuron loss. **Why this is a sourcing signal — indication expansion.** The TDP-43 platform's original addressable market was ALS/FTD (~30K ALS + ~50K FTD in the US; premium orphan pricing). This paper reframes the addressable market — if TDP-43 LOF is upstream of tau proteotoxic conversion in AD/ADRD (in the ~40% of AD patients who have TDP-43 co-pathology at autopsy per the LATE-NC literature), then a TDP-43-splicing-restoring or cryptic-exon-blocking asset acquires **Alzheimer's-scale market access** — a 5–10-million-patient reframe. Same molecule, dramatically bigger opportunity. **Two therapeutic entry points come off this paper.** Entry point 1 — restore TDP-43 splicing function; existing ASO camp already anchors this (Ionis-Amylyx-QurAlis-Trace Neuroscience STMN2 program). Entry point 2 — the newer nomination — block caspase-3-mediated tau endoproteolysis specifically in TDP-43-deficient neurons. Global caspase-3 inhibition breaks physiological apoptosis, but a *context-specific* caspase-3 vulnerability, addressable via allosteric or substrate-competitive tau-cleavage inhibitors, is a different medchem conversation. Novel target. Fresh IP. **IP + lab-status read.** Wong is coinventor on **WO2016205615A1** ("TDP-43 in degenerative disease") filed 2016 out of JHU — foundational patent family covering early TDP-43-cryptic-splicing detection + modulation. Wong is speaking at the **ALS Drug Development Summit 2026** — industry-facing programming. Grant portfolio: continuous NIH R01 funding through R01NS095969 for over a decade; currently also holds R33/R61NS115161 (translational-stage NIH mechanisms) + an ADDF grant — grant portfolio pattern of a lab *pushing toward* translation. **The 2026-07-04 COI declares "no competing interests."** That is the interesting negative signal. Wong's baseline patent estate is not zero; a "no competing interests" declaration on a 2026 paper opening a fresh indication-expansion mechanism suggests either (a) the specific caspase-3-tau-cleavage-in-TDP-43-context mechanism disclosed here is not yet wrapped into an active licensing deal, or (b) active licenses do not cover this specific mechanism claim. Either way, this is the moment where JHTV would be talking to Wong about a *new* provisional on the caspase-3-tau-cleavage-in-TDP-43-context mechanism. Exactly the JHTV timing Blackbird wants to catch. **Co-first-author signal:** Meghraj S. Baghel + Grace D. Burns are cofirst — Baghel is the more senior postdoctoral hand, watch whether he launches an independent group or moves into industry. **Competitive positioning.** Mol Neurodegener is not Nature Neuroscience but is the go-to venue for AD-mechanism papers with a translational bent; every neurology BD team at Denali, Alector, Prothena, Voyager, and Genentech Neuro scans it. Ionis-Amylyx-QurAlis-Trace will read this as an expansion opportunity for their existing splicing-restoration assets. **The JHTV window on the Wong lab's next filing is narrower than it looks — industry attention will compress the timing on any provisional-to-licensing conversation.** **Actions.** **Esther** — routine JHTV touchpoint on Wong lab intellectual-property status. Specific ask: is there a fresh provisional on the caspase-3-tau-cleavage-in-TDP-43-deficient-neurons mechanism? Current field-of-use structure of WO2016205615A1 — is it fully licensed, to whom, in what field? **Hemaka** — courtesy touch to Wong via Neuroscience/Pathology network. Pitch: "the AD/ADRD indication expansion is real, we saw the paper, what would a partnering conversation look like for Blackbird — either sourcing partner on JHTV IP or company-building around the caspase-3-tau axis." Wong is already industry-facing (ALS DD Summit), so warm not cold. **Avi** — molecular-discovery scoping on caspase-3 substrate-competitive inhibitors. Is context-specific caspase-3 modulation addressable via substrate-competitive/allosteric approaches a 2-year chemistry project or a 5-year discovery-biology problem? That answer determines whether Blackbird can run a GPR52-style 9-month medchem-to-lead campaign here or needs to underwrite a longer discovery arc. **Eddie** — file this alongside the Lieber/Third Rock schizophrenia GPR52 NewCo playbook. Both are Johns Hopkins-anchored translational neuroscience programs with a defined lead-to-NewCo pathway — different indications (AD/ADRD vs schizophrenia), different modalities (splicing correction or caspase-3-context modulation vs GPCR agonism), but a comparable license-to-NewCo shape. **Portfolio-memory addendum recommendation:** consider adding a "Wong lab TDP-43 platform" entry to the §2b JHU-adjacent Baltimore ecosystem section of blackbird-portfolio.md as an active sourcing lead — foundational IP (WO2016205615A1), translational-stage NIH funding (R33/R61NS115161 + ADDF), industry-facing (ALS DD Summit 2026), new mechanism disclosure (this paper), no active spinout yet. Paper link: https://doi.org/10.1186/s13024-026-00968-8

Sourcing Radar — Wilmer / JHU BME 9-month GLP tox preprint (Popel + Green + Pandey + Campochiaro) advances AsclepiX AXT107 suprachoroidal microparticle formulation to IND-enabling maturity, and quietly discloses a stealth new JHU-adjacent spinout called OptaNova Pharma (JJG + ASP + NBP co-founders, Wilmington DE LLC, no other public footprint) — the third company from this founder cluster after AsclepiX (2014, Phase 1/2a AXT107 in wet AMD) and Terebra Therapeutics (JHU-licensed oncolytic peptide NF27 for rectal cancer, stealth, TEDCO MII-funded); plus a JHU Pathology + Wilmer Meibomian-gland sebaceous-carcinoma paper naming MYC as the functional driver

Mon, 06 Jul 2026 12:00:00 +0000

Sourcing Radar for Monday, July 6, 2026. One institutional-intelligence lead + one Wilmer orphan-oncology watch-item. Lead. Mirando AC (first; JHU BME + AsclepiX Therapeutics), Lima e Silva R, Shen J, Robinson TJ, Green JJ (JHU BME + INBT; co-founder AsclepiX + OptaNova Pharma), Campochiaro PA (JHU Wilmer + BME; consultant AsclepiX + Cove + Genentech + others), Popel AS (senior co-anchor; JHU BME + Medicine + Oncology; co-founder AsclepiX + OptaNova Pharma + Terebra Therapeutics), Pandey NB (senior, corresponding; JHU BME part-time faculty + AsclepiX VP R&D; co-founder OptaNova Pharma + Terebra Therapeutics), "Suprachoroidal Delivery of Anti-Angiogenic Peptide Microparticles Enables Sustained Activity with Favorable Ocular Safety," bioRxiv v1 2026-07-05, DOI 10.64898/2026.06.30.735614. Funding: NIH NEI R01EY028996-06. Competing interest: **ASP + JJG cofounders of AsclepiX Therapeutics; ACM + TJR + NBP employees at time of studies; ACM + TJR + NBP inventors on pending patent 19/105,901; JJG + ASP + NBP cofounders of OptaNova Pharma; ACM + NBP + ASP cofounders of Terebra Therapeutics, LLC (unrelated to this work).** The asset: AXT107 (gersizangitide) is a collagen-IV-derived anti-angiogenic peptide disrupting α5β1 integrin + potentiating Ang2/Tie2 signaling. Discovered in the Popel Systems Biology Lab at JHU BME mid-2010s + licensed to AsclepiX 2014. Currently in DISCOVER Phase 1/2a suprachoroidal-injection trial in wet AMD (15 pts, 125/250/500 μg, enrollment complete May 2024, topline originally expected Q2 2025 per company disclosure — no shipped topline yet in public tracking, either delayed or embargoed). AsclepiX raised $10M July 2023 for Phase 1/2a. What this paper adds: **9-month GLP toxicology in Göttingen minipigs** supporting suprachoroidal microparticle formulation. Rat laser-induced CNV model — ~60% reduction of neovascular area at MP-AXT107 dose vs vehicle. 9-month refrigerated microparticle stability. Minipig 9-month single-dose safety — no treatment-related ocular findings, transient IOP + mild inflammation resolving quickly, NOAEL at top dose 1.25 mg/eye. Bioanalysis — persistent AXT107 in choroid/RPE + scleral tissues, no systemic exposure. IND-enabling regulatory-quality data — the kind of package a strategic acquirer or Series B investor wants to see. Small scientifically-honest admission — "the original, soluble AXT107 formulation was ineffective at inhibiting laser-induced CNV in our rat model and was consequently reformulated as microparticles." Candid disclosure of scientific pivot — first-gen soluble AXT107 didn't work in standard preclinical AMD model + entire microparticle reformulation is what enabled the drug to move forward. **The paper's contribution to Blackbird is in the COI disclosure, not the asset story.** Three companies named. (1) AsclepiX Therapeutics — Popel + Green cofounders 2014, well-known, not new intel. (2) **Terebra Therapeutics, LLC** — Mirando + Pandey + Popel cofounders, explicitly "unrelated to this work." Publicly identifiable: stealth Hopkins-licensed oncology company developing **oncolytic peptide NF27 for rectal cancer**, TEDCO Maryland Innovation Initiative-funded, preprint 10.1101/2025.08.17.668569 (2025-08-17). (3) **OptaNova Pharma** — Green + Popel + Pandey cofounders. **New intel.** Public footprint = only a Delaware LLC filing (Wilmington DE) on Bizapedia. No press, no website, no LinkedIn, no PitchBook, no trials, no patents. Stealth-stealth. COI is first place OptaNova has surfaced publicly. **Founder-cluster pattern.** Popel + Green + Pandey at Wilmer + JHU BME running three companies concurrently across three therapeutic areas: (a) AsclepiX = ocular anti-angiogenic peptides; (b) Terebra = oncolytic peptides for cancer; (c) OptaNova = unknown, almost certainly peptide-therapeutics given the founder overlap + lab's 15-year technology stack of mining human proteins for anti-angiogenic + oncolytic peptide sequences. **Serial peptide-therapeutics spinout factory.** Spinout-factory pattern Blackbird should be underwriting rather than watching. Aletira, Kalhor 3DEEP, Revivify Innovations, and now this Popel-Green-Pandey cluster are all Wilmer + JHU BME + Center for Nanomedicine adjacent — defensible institutional concentration mapping onto the Wilmer-cluster thesis anchoring Aletira. **Actions.** Esther — JHTV Wilmer + BME + Center for Nanomedicine portfolio-manager touchpoint this week: pending patent 19/105,901 status; AsclepiX license commercial-pipeline update; **most importantly — is JHTV managing an OptaNova license, or is OptaNova operating on inventor-owned IP that predates a Hopkins disclosure obligation.** That's the specific intelligence gap. Hemaka — courtesy touch to Popel or Pandey via BME network, pitch: "impressive that you're running three companies; where does next platform anchor sit; does Blackbird BioVentures have way to lean in on OptaNova as Seed." Eddie — file Popel-Green-Pandey cluster in Wilmer-cluster sourcing map anchoring Aletira. Bridget — if OptaNova has bench-space needs, Blackbird BioHub has capacity (soft-touch inbound). **Portfolio-memory update recommended this run:** add Popel-Green-Pandey serial-founder cluster + OptaNova Pharma + Terebra Therapeutics to Wilmer-cluster JHU-adjacent Baltimore ecosystem section of blackbird-portfolio.md. Pick 2 (watch-item). Boyack I / Berlied A / Lee S-C (JHU Pathology) / Campbell AA (JHU Wilmer) / Eberhart CG (senior co-corresponding; JHU Pathology + Ophthalmology) / Peterson C (senior co-corresponding; Tufts VetMed + Tufts SoM Ophthalmology), "MYC is functionally required in both normal and neoplastic Meibomian glands," *Am J Pathol* 2026-07-02, DOI 10.1016/j.ajpath.2026.05.013 (PMID 42392260). Peer-reviewed. No COI or patent disclosure surfaced. Ocular adnexal sebaceous carcinoma (SebCA) is aggressive high-mortality eyelid tumor arising from Meibomian glands — lacked cleanly-nominated functional oncogenic driver until this paper. MG-specific MYC inhibition in mice reduces cytoplasmic volume + proliferation + lipid droplet formation + increases apoptosis. Three MYC pharmacological tool compounds (MYCMI6 = MYC-MAX heterodimer disruptor, MYCi361 = MYC transcriptional-activity inhibitor, 10074-G5 = MYC-MAX binder) all suppress SebCA viability + clonogenicity; MYCMI6 gives cleanest transcriptional readout on canonical MYC targets. **Watch, not lead.** MYC notoriously undruggable for 30 years; three cited compounds are research-grade not clinical; no patent/company disclosure; small addressable population. Becomes portfolio-adjacent if a MYC-degrader (PROTAC) campaign for ocular adnexal SebCA emerges at Wilmer in next 12-18 months (Wilmer clinical franchise + orphan-tumor + degrader chemistry = Blackbird-plausible package). No routing action this week; add JHU Pathology + Wilmer sebaceous-carcinoma work to Wilmer memory map. **Editorial framing.** Thin Monday for a standalone paper hero. Compounding institutional-intelligence signal from Popel-Green-Pandey COI disclosure (three companies, one stealth-stealth) is more Blackbird-relevant than a marginal watch-item. OptaNova Pharma is the specific handle Esther + Hemaka should chase this week. Paper links: https://www.biorxiv.org/content/10.64898/2026.06.30.735614v1 ; https://doi.org/10.1016/j.ajpath.2026.05.013

Sourcing Radar — Reza Kalhor's JHU BME 3DEEP paper in Cell (July 1) drops a DNase-based tissue-clearing method that pushes in-situ spatial transcriptomics to 400-micron depth with a fresh JHTV patent application and no company yet — reframing a paper this desk dropped Thursday from "hair follicle biology" to what it actually is, a spatial-omics platform pivot; plus the Siliciano lab PNAS paper defining Inhibitory Potential as a rebound-competency biomarker layered on top of the JHU-licensed IPDA franchise inside Accelevir Diagnostics

Sun, 05 Jul 2026 12:00:00 +0000

Sourcing Radar for Sunday, July 5, 2026. One platform lead + one franchise-extension signal + one biophysics watch — appropriate for a thin US-holiday-suppressed week. Lead. Asami S (first; JHU BME + Center for Epigenetics), Yin C, Fan J, Garza LA (JHU Dermatology), Kalhor R (senior, corresponding; JHU BME + Center for Epigenetics + Medicine + Neuroscience + Genetic Medicine + Molecular Biology & Genetics + Kavli Neuroscience Discovery Institute), "Four-dimensional molecular mapping from a spatial snapshot reveals the dynamics of hair follicle organogenesis," *Cell* 2026-07-01, DOI 10.1016/j.cell.2026.06.014, PMID 42385701. Competing interest: **S.A. and R.K. are co-inventors of a patent application based on the methods in this study.** No assignee-company disclosure; likely JHTV-managed. **Reframing note** — this paper was surfaced on the 2026-07-02 candidate funnel and dropped as "hair follicle biology, not core Blackbird thesis." That framing was wrong. Hair follicle is the demo; the product is a spatial-omics platform. **3DEEP** = 3D DNase-Enhanced Expression Profiling. Enzymatic DNase clearing removes the dominant scattering/crowding contributor to in-situ transcriptomics depth. Achieves **spatial transcriptomics in sections up to 400 μm thick** with only 2-fold detection variation between center and surface. Every commercial spatial-transcriptomics vendor — 10x Xenium, Vizgen MERSCOPE, Bruker/CosMx — is limited to ~10-30 μm optical section. Depth extension of ~15-40x is a genuine differentiator. Authors write explicitly that 3DEEP "can be integrated with other in-situ transcriptomic approaches and tissue expansion where it may also yield gains in performance" — bolt-on positioning that spatial-omics customers want. Hair follicle demo yields three distinct organogenesis stages, 4D map from static snapshot via molecular pseudotime, IRS + hair shaft concurrent emergence between pseudotime 0.35-0.55 (contradicts histology literature), stem cell stratification established before bulb formation. Foxn1-null nude mice show 21% pseudotime delay + 26% fewer hair shaft cells — assay detects sub-clinical developmental derangement preceding overt structural defect. Authors nominate "kidney nephrons, islets of Langerhans, and lung alveoli" as next application targets — assays big-pharma diabetes + nephrology drug discovery pays for. Commercial angles: (1) license 3DEEP to Bruker/CosMx, 10x, or Vizgen as sample-prep + assay bolt-on; (2) CRO/screening business offering 3DEEP-enabled deep-tissue spatial transcriptomics to pharma; (3) combine DNase-clearing + pseudotime-from-snapshot software layer into integrated developmental-phenotyping platform. Blackbird spinout candidate should be (1)+(3) — software layer is harder-to-copy piece. Kalhor is a serial method-developer (prior MEMOIR/GESTALT-family molecular recording + single-cell lineage tracing); prior tech moved through industrial partnerships not spinouts — exactly the negotiation posture where Blackbird incubation model creates leverage. **Actions — Esther + Hemaka + Avi.** Esther on JHTV BME portfolio manager this week: patent status on 3DEEP application, license availability, whether Kalhor is open to spinout structure or already spoken for. Hemaka on courtesy touch to Kalhor. Avi on scoping software layer (pseudotime-from-snapshot is where the real IP moat sits). Competitive urgency real — Cell publication is the trigger for spatial-omics vendor BD teams to move; whichever accelerator gets to Kalhor first sets the negotiation frame. Pick 2 (franchise-extension signal). Garcia MA (first; now AstraZeneca, disclosed), ..., Cohn LB (Fred Hutch + UW; senior co-corresponding), Simonetti FR (Gilead consulting; disclosed), Siliciano RF (JHU Medicine + HHMI; senior), Siliciano JD (senior, corresponding; JHU Medicine), "Inhibitory potential of autologous neutralizing antibodies sets quantitative limits on the rebound-competent HIV-1 reservoir," *PNAS* 123(27):e2608337123, Epub 2026-07-02, DOI 10.1073/pnas.2608337123, PMID 42391404. Competing interest: HIV-1 IPDA is patented (PCT/US16/28822) through JHU with R.F. Siliciano as inventor; **Accelevir Diagnostics holds exclusive licensing rights; Siliciano holds no equity.** 13 ART-interruption participants; rebound viruses are genetically identical/near-identical to circulating reservoir proviruses. New quantitative biomarker — **Inhibitory Potential (IP)** = log-reduction in single-round infection at physiologic IgG from patient's own contemporary plasma. Rebound viruses show IP 0.5-2.8 logs (up to 631× suppression) — insufficient. Reservoir variants that stay dormant show IP 0.4-8.2 logs. As autologous NAb potency wanes on ART, previously-neutralized reservoir variants regain infectivity + become rebound-competent. Commercial angle — **franchise extension, not spinout.** IPDA already licensed exclusively to Accelevir Diagnostics (Baltimore-anchored HIV-cure-companion Dx). Inhibitory Potential is the natural franchise extension — Ph2/3 HIV cure trial stratifier layered on IPDA reservoir-quantification workflow. HIV cure trials (Merck islatravir, Gilead lenacapavir combos, ViiV cabotegravir, academic ATI consortium) burn tens of millions per protocol on treatment interruptions where >50% rebound. Validated IP threshold has real pull. **Portfolio-memory action** — Accelevir added to JHU-adjacent Baltimore ecosystem map. Ecosystem signal — first author now at AstraZeneca; Simonetti gets Gilead consulting. Big pharma pulling talent + consulting relationships out of Siliciano lab around biomarker franchise. Pick 3 (biophysics watch-item). Yang F, Moulick R, Wang C, Rodgers ML (NIH IRP), Woodson SA (JHU T.C. Jenkins Biophysics; senior co-anchor), Zhang Y (senior, corresponding; JHU), "Analytical framework for molecular dwell times in biomolecular condensates," bioRxiv v1 2026-07-03, DOI 10.64898/2026.06.29.735418. Theoretical framework — condensate dwell-time distributions follow power-law × exponential form; exponent −1.5 = diffusion-limited escape, exponent −0.5 = interfacial-barrier-limited escape. Validated via single-molecule RNA measurements in in-vitro condensates. Design-time diagnostic for every condensate-drug-discovery company (Dewpoint, Nereid, Faze/Novartis, Transition Bio) — tells medicinal chemist which lever to pull (viscosity/composition for diffusion-limited, surface tension for interfacial-limited). Preprint only, no COI/patent. Pre-IP territory. Watch-item only. Avi — add Yaojun Zhang lab (JHU) to condensate-biophysics tracking list. **Editorial framing** — thin flow week (US July 4 holiday + weekend suppressed JHU/UMB/Lieber press output + slowed PubMed indexing volume). Kalhor 3DEEP is the lead if platform reframe holds. Siliciano IP is franchise-extension signal on Accelevir landscape. Woodson/Zhang is biophysics watch-item. No UMB-primary or Lieber-primary translational asset story this cycle. Not every week produces a fresh spinout candidate; forcing one on a thin week produces the wrong pick. Paper links: https://www.cell.com/cell/fulltext/S0092-8674(26)00702-6 ; https://www.pnas.org/doi/10.1073/pnas.2608337123 ; https://www.biorxiv.org/content/10.64898/2026.06.29.735418v1

Portfolio Watch — Roche divarasib beats Amgen Lumakras + BMS Krazati head-to-head in Phase 3 KRAS G12C NSCLC (aSKY KRAS-adjacent read); Abivax obefazimod Phase 3 UC malignancy signal cleared (Slusher GCPII IBD threat + RNA-upregulator platform validation); Vijay Kumar exit + Karim Mikhail acting the CBER-OTP office (Aletira regulatory tailwind); United Therapeutics buys Thymmune for $140M upfront + $160M earn-outs (preclinical platform comp); Vertex Casgevy expanded to age-2+ (gene-therapy regulatory softening); IPO window open + XBI at five-year high (financing backdrop)

Sun, 05 Jul 2026 13:00:00 +0000

Portfolio Watch for the week ending Sunday, July 5, 2026. Six items move the thesis this week. Two data readouts (Roche divarasib + Abivax obefazimod) + four ambient signals (CBER-OTP transition, UTHR-Thymmune deal, Casgevy pediatric expansion, IPO window). **Item 1 — Roche divarasib Krascendo-1 (2026-07-02).** First Phase 3 head-to-head between a next-gen KRAS G12C inhibitor and both approved competitors (Amgen Lumakras/sotorasib + BMS Krazati/adagrasib). Divarasib delivered statistically significant OS + PFS wins in previously-treated KRAS G12C NSCLC. Next Roche readout is Krascendo-2 (divarasib + Keytruda first-line) — the $6B question for the class. **aSKY read** — validation with caveat. aSKY is KRAS-adjacent oncology (UMB Krupnick + AVS Bio antibody engineering with KRAS frame per LinkedIn late 2025). Validation — KRAS-directed oncology is live differentiable class, big-pharma appetite intact. Caveat — ceiling for direct G12C TKI approaches just got higher. Any KRAS-adjacent program must position complementary to (not competing with) next-gen TKI class — combination synergy, resistance-mechanism coverage, or non-G12C-mutant tumors. Krupnick's lung-cancer NK-cell background is naturally immune-adjacency, not TKI-adjacency — that's where aSKY positioning holds up cleanly. **Actions** — Eddie/Hemaka sanity-check aSKY narrative continues immune-adjacency + KRAS-pathway biology, not TKI-adjacent; if AVS Bio partnership deliverables include KRAS-mutant subtypes or resistance coverage, that's the differentiation. Watch for Amgen/BMS defensive acquisitions/partnering around next-gen KRAS in next 2 quarters. **Item 2 — Abivax obefazimod Phase 3 UC malignancy cleared (2026-07-01).** Supplementary ABTECT analysis on ~1,700 patient-years — cancer rates at background UC-population levels. Signal resolved; Crohn's expansion signaled. Obefazimod is oral small molecule upregulating miR-124 → downregulates pro-inflammatory genes — opposite direction of standard playbook (silence pathological RNA). Most clinically advanced RNA-upregulator small-molecule in inflammation. **Portfolio read #1 (threat) — Slusher gut-restricted GCPII IBD program.** Cleaner-safety oral small molecule with prevalent-disease commercial ambition ($10B+ Crohn's+UC market) is now de-risked. Doesn't kill GCPII (different mechanism — glutamate-carboxypeptidase-II inflammation + enteric-nervous-system pain vs miR-124 pro-inflammatory-gene modulation) but ceiling for oral small-molecule IBD entrants is harder to cross. **Slusher differentiation must be the pain axis** — GCPII's bimodal effect on inflammation AND enteric neuropathic pain is what obefazimod does not touch. Pain is real + enormous IBD unmet need. **Portfolio read #2 (validation) — RNA translation-activator platform.** Abivax cleared safety on a Phase 3 RNA-modulating small molecule that upregulates a target — closest live proof point that "increase, not silence" can produce a pill-form blockbuster in prevalent inflammatory disease. Mechanistic direction different (obefazimod = miRNA induction to suppress inflammation; Blackbird platform = translation activation to raise haploinsufficient protein) but category is same — RNA-directed small molecules that increase rather than silence. **Avi** — cite Abivax as market-validation data point on next outside pitch. Camp4 also validated. With Ionis Tryngolza label expansion (2026-06-24) + Abivax obefazimod safety (this week) + Remix-Passage reverse merger (last week) = **three RNA-modulator commercial validation points in 8 days.** Category re-rating. **Item 3 — FDA CBER-OTP Vijay Kumar exit (2026-06-30).** Karim Mikhail (acting CBER Director) overseeing the office directly during internal/external search. BioPharma Dive frames this as reinforcing post-Makary pivot toward more flexibility for rare-disease/genetic-medicine sponsors — office recently reversed course on Uniqure, Regenxbio, Atara. **Aletira touch** — this is the office reviewing Aletira's future IND on hereditary hearing-loss AAV/SELEXON. Direct tailwind. Modestly favorable for gene-therapy M&A appetite that could ultimately buy Aletira (Lilly-Akouos-style acquirers watch CBER tone for exit-path clarity). Combined with last-week Sangamo Chapter 11 + Passage Bio reverse merger — AAV field being simultaneously repriced downward on standalone-company valuations AND re-emphasized by regulator posture. Market separating strategic acquirers (Lilly, Roche, UTHR) from standalone AAV companies (Sangamo, Passage). Aletira's preclinical-AAV-with-partnering-exit model is exactly right structure for this repricing. Eddie tracks CBER-OTP search process. **Item 4 — UTHR acquires Thymmune $140M upfront + $160M earn-outs (2026-07-02).** Preclinical iPSC thymic-cell-therapy platform (congenital athymia + post-transplant tolerance + immune reconstitution). Back-loaded milestone structure = norm for pre-clinical platform deals; modest upfront, earn-outs on preclinical → IND → Ph1 milestones. **Term-sheet comp for Aletira, RNA translation-activator platform, future Slusher spinout** — models on Thymmune-style structure, not $500M-all-upfront numbers like Apogee. UTHR joins Lilly, Roche, GSK as active mid-cap-to-mega-cap acquirers. Strategic-acquirer roster broadening. Eddie files in term-sheet-comp library. **Item 5 — Vertex Casgevy expanded to age-2+ SCD/TDT (2026-07-01).** First CRISPR-edited cell therapy pediatric label extension. FDA signaling comfort with pediatric benefit-risk on genetic medicines. Consistent with CBER-OTP tone shift. General regulatory-tone tailwind for Aletira (hereditary hearing loss = pediatric-onset). **Item 6 — Financing environment.** BIO 2026 read = "window is open." ~12 biotech IPOs likely Q3 2026. Kardigan (BMS/MyoKardia CV precision-med spinout) priced $400M IPO. YTD 11/13 biotech IPOs raised >$250M — highest quality bar since 2018. XBI +30% YTD, +19% in June, approaching Feb-2021 peak. **Filter — asset-centric, Ph2+, experienced management. Not preclinical platform stories.** Aletira + RNA translation-activator platform will face "asset-centric Ph2+" filter — accelerate clinical-entry timelines, don't raise on preclinical platform slides alone. Neuropsych GPR52 NewCo, if DC lands late 2026 as planned, has legitimate Series A/B window through Q4. **Cross-cutting frames.** (1) Lilly pattern continues — no new Lilly moves (integrating Sangamo/Verve/Akouos) but CBER-OTP tone shift + UTHR-Thymmune deal reinforce strategic-acquirer roster broadening. GSK inside Nuvalent, UTHR inside Thymmune, Roche doubling on KRAS. Lilly remains single most active gene-therapy strategic — Eddie continues Lilly BD relationship-building AND UTHR relationship. (2) RNA-modulator category continues systematic validation — 3 commercial validation points in 8 days. Blackbird translation-activator platform positioned into that tailwind if it moves this window. **Quiet week on direct portfolio front** — nothing from 1104health, aSKY, Aletira, GPR52 NewCo directly or from Baltimore ecosystem (no BioBuzz-level Baltimore deal). Active on surrounding landscape. Press links: https://www.roche.com/media/releases/med-cor-2026-07-02 ; https://www.biopharmadive.com/news/roche-kras-lung-cancer-data-head-to-head/824344/ ; https://www.biopharmadive.com/news/abivax-obefazimod-Phase3-safety-data-cancer-ulcerative-colitis/824196/ ; https://www.biopharmadive.com/news/top-fda-gene-cell-therapy-regulator-step-down/824093/ ; https://ir.unither.com/press-releases/2026/07-02-2026-113022825 ; https://news.vrtx.com/news-releases/news-release-details/vertex-announces-us-fda-approval-expanded-use-casgevyr-treatment ; https://www.biopharmadive.com/news/bio-2026-ipo-biotech-performance-predictions/823474/

Sourcing Radar — Jiou Wang's Packard-Center-funded eLife paper delivers a minimally-disruptive Anap noncanonical-amino-acid live-cell imaging platform for TDP-43 and stress granules, paired with the same-day Nature Aging XL20 first-in-class TDP-43 conserved-region small molecule that extends ALS-mouse survival and names Pan P. Li's JHU Psychiatry lab as JHU intersect + a Chatterjee/Schlosser JHU Bloomberg Sci Transl Med proteome-wide association study that nominates LILRB1 and SIRPα (both already drugged in oncology) as new causal Alzheimer's-disease targets

Sat, 04 Jul 2026 12:00:00 +0000

Sourcing Radar for Saturday, July 4, 2026. Three-pick radar. Lead. Chen H, Wang Ha, Lu Y-N, Chen P, Zheng Z, Zhang T, Wang J (senior, corresponding; JHU Biochem & Mol Biology, Bloomberg SPH + JHU Neuroscience, SOM), "Non-canonical amino acid incorporation enables minimally disruptive labeling of stress granule and TDP-43 proteinopathy," *eLife* 2026-07-03, DOI 10.7554/eLife.109452 (PMID 42397263). Packard Center for ALS Research at JHU + NIH funding; no competing interests declared. Genetic-code-expansion + polarity-sensitive fluorescent noncanonical amino acid **Anap** incorporated at G3BP1 F337 and TDP-43 V100 via amber-stop-codon reassignment + orthogonal aminoacyl-tRNA synthetase/tRNA pair — ~200 Da, intrinsically fluorescent, functionally invisible when placed at rational sites. Anap-labeled G3BP1 shows ~53% FRAP mobile fraction in stress granules; GFP-G3BP1 shows ~33% — GFP measurably rigidifies condensates. Anap-labeled TDP-43 stays diffuse and nuclear at baseline, forms liquid-like nuclear puncta only under mild stress, ~45% FRAP recovery; YFP-TDP-43 forms aberrant nuclear puncta at baseline with solid-like 22% recovery. Anap incorporation preserves TDP-43 splicing activity; YFP tagging does not. Detects native nuclear G3BP1 pools antibody methods miss. **Commercial angle** — screening platform, not therapeutic asset. Two immediate use-cases: (1) high-content live-cell screens for compounds that dissolve TDP-43 aggregates without disturbing splicing (exactly the profile any TDP-43-directed ALS drug needs); (2) MOA deconvolution for pipeline programs (XL20 below, Ionis/Biogen ATXN2 ASOs). Generalizable to FUS, hnRNPA1/A2. IP path — GCE-Anap + site-selection rules are patentable know-how; no COI/patent disclosure in eLife but JHTV filing status warrants confirmation. Blackbird routing — **Avi + Hemaka via Packard Center intro**; Packard is Blackbird's most naturally-aligned JHU ALS/FTD partner. Pick 2 (landscape wave, Arizona-primary IP). Gao J / Shukla D / ... / Tang F (JHU Psychiatry) / Li PP (JHU Psychiatry & Behavioral Sciences, Div of Neurobiology) / Wang X (senior, corresponding; UArizona College of Pharmacy Pharm & Tox), "Therapeutic targeting of the conserved region within the low-complexity domain of TDP-43 is neuroprotective and extends survival in ALS mice," *Nature Aging* 2026-07-03, DOI 10.1038/s43587-026-01166-3 (PMID 42399370). Structure-based virtual screening identified brain-penetrant small molecule **XL20** binding TDP-43's conserved region (CR; residues 320-340 in low-complexity C-terminal domain). Micromolar SPR affinity — CR deletion abolishes binding, Trp334 mutation markedly reduces signal. Efficacy stack: rat primary neurons rescued at 6.25 μM; iPSC motor neurons carrying p.Q331K ALS mutation restored to isogenic-control spine density, cytoplasmic TDP-43, mitochondrial length + membrane potential at 20 μM x 7 days; **TDP-43 p.A315T transgenic ALS mice** — oral XL20 extends survival meaningfully in males, eliminates hindlimb clasping at day 70, restores rotarod to non-transgenic-littermate levels, reduces motor neuron loss + muscle atrophy. **Critical control** — XL20 does NOT affect TDP-43-dependent RNA splicing (ITPR3 splicing preserved) — the fidelity signal that distinguishes CR-targeting from ASO knockdown approaches. Mechanism — CR engagement suppresses TDP-43 mitochondrial localization + prolonged "kiss-and-run" toxicity; XL20 restores basal + maximal respiratory capacity + Complex I. **IP structure** — Wang X. (Arizona) senior + majority of coauthors at UArizona College of Pharmacy — **primary IP at Arizona**, not JHU. JHU contribution = **Pan P. Li lab + Fan Tang** in Psychiatry Neurobiology; based on Li publication history her group likely contributed CR-mediated LLPS + mitochondrial-localization validation. XL20 is not a Blackbird license candidate (Arizona-owned). But Nature Aging landing marks the category — small-molecule CR-targeting for TDP-43 proteinopathies — that will attract IND-stage capital in next 12-18 months. Adjacent JHU-owned TDP-43 chemistry, screening hits, structural biology tools wants to move now. **Esther Park action** — Fellow-style scoping conversation with Pan Li lab in next diligence cycle. Adjacent landscape — Biogen/Ionis tofersen (SOD1), Denali BIIB105 (Ataxin-2), Neuvivo, Insilico INS018_055, academic TDP-43 programs at UCSF/Yale/Stanford. XL20 differentiation — small molecule + brain-penetrant + splicing-sparing + survival extension in standard TDP-43 model. Legitimate first-in-class TDP-43 small-molecule candidate. Pick 3 (landscape signal, portfolio-adjacent — not sourceable). Walker KA / Blew C / ... / Chatterjee N (senior; JHU Bloomberg SPH Biostatistics) / Schlosser P (senior, corresponding; JHU Bloomberg SPH Epidemiology + Univ. Freiburg), "Alzheimer's disease proteome-wide association study implicates adaptive immunity and identifies risk genes LILRB1 and SIRPA," *Sci Transl Med* 18(856):eadx4852, 2026-07-01, DOI 10.1126/scitranslmed.adx4852 (PMID 42384774). PWAS integrating plasma cis-pQTL data (1,348 European-American + 1,385 African-American genetically-determined protein models) with AD dementia GWAS + colocalization + SMR causal inference. 18 candidate genes in European-American cohort; 13 supported by colocalization/SMR. Four new to AD — CD55, **LILRB1**, SCARA5, **SIRPA**. LILRB1 + SIRPα strongest mechanistic linkage. **Why Blackbird cares** — both already drugged in oncology. SIRPα is the receptor pair of CD47 (Gilead/Forty Seven magrolimab, Trillium/Pfizer TTI-621/622, Arch Oncology, multiple SIRPα-Fc fusions). LILRB1 has clinical programs at Merck, Immune-Onc, NextCure. This is repurposing thesis territory — two peripherally-druggable AD targets with pre-baked drug discovery. Blackbird does not have an AD program; take-home is that AD-as-immune-disease target list is expanding + new candidates carry oncology-derived chemistry already in clinic. Landscape signal for repurposing-thesis future play. JHU role is statistical (Chatterjee + Schlosser Bloomberg SPH), not therapeutic — **not JHU-sourceable directly**. Blackbird sourcing signal — TDP-43 concentration at JHU. Two same-day papers converge on TDP-43 at JHU: eLife JHU-primary imaging platform (Jiou Wang, Packard) + Nature Aging Arizona-primary compound paper (Wang X. / Li P. P., JHU co-contribution). JHU TDP-43 ecosystem thicker than portfolio memory currently reflects. **Portfolio-memory expansion recommended**: Jiou Wang lab (Biochem BSPH + Neuroscience SOM; stress-granule + TDP-43 + GCE-Anap imaging); Pan P. Li lab (Psychiatry Neurobiology; TDP-43 CR + LLPS + mitochondrial toxicity + iPSC motor neurons); Packard Center for ALS Research at JHU (institutional funder + Blackbird conversation channel via Tremblay's operator network). Paper links: https://elifesciences.org/articles/109452 ; https://www.nature.com/articles/s43587-026-01166-3 ; https://www.science.org/doi/10.1126/scitranslmed.adx4852

Sourcing Radar — Lieber Weinberger-group Mol Psychiatry paper finds two novel divergent transcripts at the BDNF locus (BDNF-DT + BDNF-AS-DT) that are activity-dependent, developmentally regulated, and genetically associated with schizophrenia — a native regulator that intersects both the Lieber GPR52 program and Blackbird's translation-activator RNA-upregulation platform

Fri, 03 Jul 2026 12:00:00 +0000

Sourcing Radar for Friday, July 3, 2026. One clear lead + one watch-item. Lead. Bach SV (Lieber), Punzi G, Smith NE, Mukherjee S, Shin JH, Chen Q (NIH-NCATS Stem Cell Translation Lab), Pertea G, Collado-Torres L, Maynard KR, Page SC, Kleinman JE, Hyde TM, Weinberger DR, Martinowich K (senior, co-corresponding; Lieber + JHU Psychiatry + Snyder Neuroscience + Kavli), Ursini G (senior, co-corresponding; Lieber + JHU Psychiatry), "BDNF-DT and BDNF-AS-DT: novel genes in the BDNF locus," *Molecular Psychiatry* 2026-06-30, DOI 10.1038/s41380-026-03695-0 (PMID 42380610). Peer-reviewed. No competing interests declared; Lieber + NIH RO1MH105592, R01MH123567, P50MH094268 + P50MH136297 funding. Re-analysis of Lieber's postmortem dorsolateral prefrontal cortex (DLPFC) RNA-seq (Jaffe 2018) identified poly-A+ reads upstream of BDNF exon I on the antisense strand not annotated in the current reference — validated by end-to-end PCR + cloning + Sanger sequencing on independent DLPFC samples. Three main transcripts of a new gene named **BDNF-DT** (BDNF divergent transcript; DT1, 2, 3) plus one **BDNF-AS-DT** readthrough transcript formed by splicing of exons from the previously-known BDNF-AS with the new BDNF-DT. BDNF-AS-DT does *not* overlap the BDNF protein-coding sequence. Three findings that matter. **Developmental trajectory** — BDNF-DT expression rises in prenatal life, decreases in the childhood-to-adolescence transition, peaks again around age 25, then declines; that pattern mirrors BDNF itself. The previously-known BDNF-AS goes the opposite direction. BDNF-DT is positively correlated with BDNF; BDNF-AS-DT is negatively correlated; BDNF-AS shows no correlation. **Activity-dependence** — depolarization of cultured human neurons increases BDNF-DT + BDNF-AS-DT expression with dynamics that follow BDNF's activity-dependent induction; CRISPR-mediated activation of BDNF in human neural progenitor cells drives BDNF-DT expression as a downstream consequence — functional coupling. **Schizophrenia genetics** — BDNF-DT is higher in schizophrenia DLPFC than in neurotypical controls; genetically-predicted lower BDNF-AS-DT (readthrough) is associated with schizophrenia risk AND with the schizophrenia-risk C allele of rs6265 (BDNF Val66Met; Val=C, Met=T). The paper offers a molecular mechanism for the rs6265 association via a downstream regulatory RNA — the C allele's schizophrenia risk may be mediated by reduced BDNF-AS-DT expression, not just by (or in addition to) the Val→Met protein change itself. REST binding — 5' region of BDNF-DT overlaps a REST binding site on BDNF; co-expression network analyses predict REST-regulated gene programs for both BDNF-DT and BDNF, positioning BDNF-DT in the REST-BDNF regulatory axis (Huntington's-disease-relevant node). Blackbird three-way thesis intersection. **(1) Lieber schizophrenia (GPR52) program** — same institution + same disease. BDNF-DT and BDNF-AS-DT are candidate pharmacodynamic biomarkers for target engagement in the GPR52 DC + patient stratification (rs6265-genotyped subgroups). Add Gianluca Ursini (senior co-corresponding + P50MH funding) to Lieber Blackbird routing list as candidate next-NewCo PI on the BDNF regulatory-RNA axis. **(2) RNA-upregulation translation-activator platform** — BDNF is on the shortlist of any RNA-upregulation platform's target map. The Bach paper adds a new regulatory-R-N-A layer (BDNF-DT + BDNF-AS-DT) that any B-D-N-F upregulation strategy has to reckon with — either as biomarkers of successful induction or as candidate targets themselves. Camp4 Therapeutics competitive signal on the B-D-N-F target-map reshuffle. **(3) REST-BDNF axis** — REST is the master transcriptional repressor of neural genes, pharmacologically underserved; a specific instance of REST-dependent regulation with a clinical genetic readout in schizophrenia is a longer-arc handle worth ecosystem tracking. Second pick (watch-item, not sourcing lead). Chen F, Saqib M, Nguyen CM, Sarver DC, Yu YE, Aja S, Seldin MM, Wong GW (senior, corresponding; Johns Hopkins University School of Medicine, Physiology + Center for Metabolism and Obesity Research), "Gene dosage imbalance disrupts systemic metabolism in the Dp16 Down syndrome mouse model," bioRxiv v3 posted 2026-07-01, DOI 10.64898/2026.01.13.699318. Preprint, not peer-reviewed; no competing interests declared. Comprehensive multi-tissue (WAT + BAT + liver + skeletal muscle + hypothalamus) transcriptomic + metabolomic + physiological phenotype of Dp16 mice both sexes. Core phenotypes — pronounced insulin resistance, glucose intolerance, impaired lipid clearance, dyslipidemia; tissue signatures — immune activation, ER + oxidative stress, fibrosis, impaired glucose + fatty acid catabolism, altered lipid + bile acid profiles, reduced mitochondrial respiration. Obesogenic diet exacerbates insulin resistance despite sex-divergent weight gain. Foundational animal-model characterization, not target-discovery. Down syndrome individuals have elevated risk of obesity, T2D, dyslipidemia, and NAFLD with limited trisomy-21-tailored therapeutic options. Wong lab (CTRP adipokine family; long JHU tenure) is the natural JHU partner for a future D-S metabolic-therapeutic collaboration if a specific triplicated-gene target emerges from follow-on work. Watch-item, not this-week's diligence action. Actions. (1) Hemaka to Lieber Blackbird-sourced-program lead on the BDNF-DT paper for the GPR52 team's biomarker + stratification thinking; add Ursini to Lieber routing list. (2) RNA-upregulation platform team reads the Bach paper end-to-end and integrates BDNF-DT + BDNF-AS-DT into their B-D-N-F program regulatory-R-N-A target map. (3) Bookmark Wong lab Dp16 metabolic phenotype dataset for future metabolic-thesis routing. Candidate funnel — 2026-06-30 to 2026-07-03. bioRxiv + medRxiv details API scanned ~450 preprint records; JHU/Lieber/UMB corresponding-institution filter yielded 6 hits — Wong Dp16 v3 CHOSEN as second-pick watch-item; Charles JHU GLM-HMM (computational neuro method — DROPPED); Casadevall JHU Bloomberg Cryptococcus (infectious disease microbiology — DROPPED); Hopkins ICU antifungal septic shock + Hopkins Bloomberg heat-wave scenario (clinical epi + public health — DROPPED); UMB Krantz solo-author schizophrenia HCAR1/HCAR2 TAD-collapse thesis (also dropped by 2026-07-02 funnel — DROPPED). PubMed 210 JHU + 85 UMB + 5 Lieber items scanned — Bach BDNF-DT *Mol Psychiatry* CHOSEN as lead (full text via r.jina.ai clean, 191k chars); Ravichandran *Neuron* NAc atlas + Schiffers *Nature* ac4C already used as 2026-07-02 picks and skipped (do not repeat within days); Ferreira *Cancer Res* CDKN2A/ARF PDAC not JHU-led (Stanford-corresponding — DROPPED); Liang *Genome Med* NPTN autism/DD not JHU-led (Leibniz + GWU corresponding — DROPPED, would have been RNA-upregulation-adjacent haploinsufficiency target if JHU-led); Rehg mPFC-LC attention (already 2026-06-28 radar — DROPPED); Punzi *Transl Psychiatry* editorial letter (not primary research — DROPPED); remaining JHU/UMB PubMed items clinical epi, device/imaging, or behavioral/public-health without therapeutic sourcing signal. Institutional press (hub.jhu.edu, hopkinsmedicine.org, JHTV, UM Ventures, libd.org) — no in-window items with fresh therapeutic sourcing angle beyond prior radar coverage. Sources that failed transiently — hub.jhu.edu/research + biorxiv.org search returned 403 Cloudflare; r.jina.ai reader-proxy resolved both. Doi.org on 10.1016/j.neuron.2026.06.002 redirected to ScienceDirect paywall but NAc atlas already covered yesterday so no full-text pull required. Editorial call. **One clear lead + one watch-item.** The BDNF-DT paper is the strongest fresh Lieber-anchored translational-genetics finding this week — a new regulatory-RNA layer at the B-D-N-F locus with a schizophrenia-genetics-linked biomarker (BDNF-AS-DT reduced with rs6265 C allele) and Camp4-competitive signal for the Blackbird translation-activator platform. Not spinout-ready in itself but hits three Blackbird theses simultaneously. The Wong lab Dp16 v3 is a real dataset from a legitimate JHU metabolic-physiology PI but with no target-declaration or IP disclosure it's a watch-item, not this week's diligence action. Paper links: https://www.nature.com/articles/s41380-026-03695-0 ; https://www.biorxiv.org/content/10.64898/2026.01.13.699318v3

Sourcing Radar — Nature ac4C paper (Wu/Oberdoerffer/Mao) unwraps a JHTV-managed LNP patent from the Mao lab covering the mRNA modification that just beat the industry-standard N1-methylpseudouridine on both translation yield and fidelity, plus a Lieber-led human nucleus accumbens spatiomolecular atlas with an Alexis Battle CellCipher disclosure

Thu, 02 Jul 2026 12:00:00 +0000

Sourcing Radar for Thursday, July 2, 2026. One dominant lead + one portfolio-adjacent second pick. Lead. Schiffers S / Nelson BW (co-first), Prigge M, Krishna S, Watkins L, Zhu Y, Tyagi N, Beiki H, Das S, Raman A, Ma J, Andresson T, Mao HQ (senior; JHU BME + Institute for NanoBioTechnology + Translational Tissue Engineering Center + Materials Science & Engineering), Wu B (senior, corresponding; JHU Biophysics & Biophysical Chemistry + Center for Cell Dynamics + Snyder Neuroscience), Oberdoerffer S (senior, corresponding; NCI/NIH IRP), "N4-Acetylcytidine enhances synthetic mRNA translation yield and fidelity," *Nature* 2026-07-01 (DOI 10.1038/s41586-026-10729-8, PMID 42386968). Peer-reviewed. **Competing interest — H.-Q.M., J.M. and Y.Z. are co-inventors of US provisional patent application no. 63/325,222 covering the LNP formulation described in this paper. The patent was filed through and managed by the Office of Johns Hopkins Technology Ventures.** The industry-standard mRNA modification N1-methylpseudouridine (m1Ψ) — the Karikó-Weissman chemistry in every FDA-authorized mRNA vaccine + every clinical-stage mRNA therapeutic — was documented by two Nature papers (Mulroney 2024, Rozman 2026) to cause +1 ribosomal frameshifting at U-rich sequences, producing truncated / neoantigenic peptides with immune responses detected in mRNA-COVID-vaccine recipients. The Schiffers/Nelson paper is the first head-to-head demonstrating N4-acetylcytidine (ac4C) beats m1Ψ on both translation yield and fidelity, in primary human MoDCs and mouse liver. Key numbers: ribosome runoff median unmodified 3.7 min, m5C 4.2 min, ac4C 6.7 min, **m1Ψ 13.7 min** — m1Ψ elongation ~2x slower than ac4C, mechanistic basis for m1Ψ's paradoxical high ribosome occupancy + low protein output (stalling → ribosome collisions → ZNF598-mediated RQC → subunit splitting + nascent peptide degradation). Fidelity: in +1FS-5×U firefly-luciferase reporters, m1Ψ produces frameshifted full-length; ac4C does not. In +1FS-5×C reporters ac4C also does not frameshift. Immune activation (IFNB1 + TNF + p-eIF2α) equivalently suppressed by ac4C and m1Ψ in THP-1 macrophages + primary MoDCs from multiple donors + MEFs. In mouse liver 24h after IV LNP-mRNA, ac4C exceeds m1Ψ on hepatic bioluminescence. Sourcing signal. Three JHU co-inventors on the JHTV-managed LNP patent (Mao/Ma/Zhu) + no visible Mao mRNA-LNP vehicle on the JHTV public portfolio (Mao's known JHTV spinouts LifeSprout + SpaceTime are tissue-engineering, not mRNA) + ac4C chemistry itself likely NCI-Oberdoerffer IP position from 2018+ Cell paper onward = **two-office diligence needed (JHTV + NCI Tech Transfer) to determine whether a Blackbird-underwritten mRNA-LNP NewCo can practice both pieces of IP under coordinated licenses.** Actions. Esther + Hemaka to JHTV BME portfolio manager on 63/325,222 status this week (priority-tier); Avi + Hemaka to NCI Tech Transfer on ac4C chemistry position (second-priority); Hemaka + Geoff Lynn courtesy touch to Mao lab on portfolio-coherence with Aletira (AAV) + Mao (LNP) as parallel Blackbird nucleic-acid-delivery modalities; Esther + Hemaka to Bin Wu group (Snyder Neuroscience + Biophysics) on SINAPs live-cell translation-imaging platform as future NewCo QC infrastructure. Competitive urgency is real — every mRNA player (Moderna, BioNTech, Arcturus, Precision Bio, Strand, CureVac, Sanofi-Translate, TriLink, Aldevron/Danaher) will be calling this week. Second pick. Ravichandran P, Bach SV, Phillips RA III, ..., Battle A, Hicks SC (co-corresponding), Maynard KR (co-corresponding; Lieber Institute for Brain Development), "Spatiomolecular mapping reveals anatomical organization of heterogeneous cell types in the human nucleus accumbens," *Neuron* 2026-07-01 (DOI 10.1016/j.neuron.2026.06.002, PMID 42385698). Paired snRNA-seq + spatial transcriptomics on postmortem human NAc. Identifies transcriptionally unique cell populations + spatial domains (SpDs) including D1 islands (discrete MSN subtypes enriched for DRD1 + OPRM1), continuous spatial gradients, evolutionary conservation, SpDs associated with psychiatric + addiction GWAS risk, spatially mapped risk-associated ligand-receptor interactions, predicted enrichment of drug-responsive transcriptional programs in specific SpDs + cell types. Blackbird portfolio angle. The nucleus accumbens is the mesolimbic dopaminergic hub; Blackbird's Lieber+Third Rock GPR52 schizophrenia program lives in this circuit. Atlas is target-discovery infrastructure for any NAc-directed therapeutic — informs candidate spatial biodistribution, target-engagement pattern, and neighborhood ligand-receptor pairs mapping to psychiatric-risk genetics. Directly relevant for the GPR52 DC's expected NAc engagement pattern. Ecosystem signal. **Alexis Battle discloses co-founder + equity holder of CellCipher.** Battle is JHU BME + Genetic Medicine + CS + Malone Center faculty; lab does predictive modeling of impact of genetic variation on cellular phenotype — computational-genomics tooling that reads as a target-discovery-services platform. CellCipher is not on the JHTV public portfolio-companies list (2026-07-02 sweep). This is the second consecutive-day competing-interest disclosure of a stealth JHU-adjacent biotech, following Revivify Innovations from 2026-07-01. Growing pattern worth naming as systematic ecosystem intelligence. Actions on second pick. Esther on CellCipher trace (website, filings, financing state, technology thesis — priority-tier ecosystem intel); Hemaka to Kristen Maynard's Lieber group for atlas-driven target-discovery collaboration channel; Hemaka + schizophrenia-NewCo team on GPR52 spatial-context read from the atlas (SpD localization + co-expression neighborhoods + risk-associated ligand-receptor pairs intersecting the GPR52 neighborhood). Candidate funnel — 2026-06-29 to 2026-07-02. bioRxiv + medRxiv details API paginated: 367 unique preprint records processed; JHU/Lieber/UMB corresponding-institution filter yielded 5 hits — Krantz UMB single-author schizophrenia HCAR1/HCAR2 TAD-collapse thesis (florid abstract from protein biochemist w/o co-authors, red-flag pattern for solo speculative computational reanalysis — DROPPED); Garg UMB EMRE-MCU biophysics (real group but MCU has pharmacology history w/o clinical success — DROPPED as basic mechanism/tool); NPTX2 v3 (Snyder, core v1 October 2025 — DROPPED); antifungal septic shock + genetic counseling videos (JHU medRxiv, clinical epi + educational — DROPPED). PubMed 100+ JHU + 6 UMB + 2 Lieber hits, therapeutic-flagged shortlist: **Schiffers/.../Mao/Wu/Oberdoerffer *Nature* CHOSEN as lead**, full text via r.jina.ai clean fetch (198k chars); **Ravichandran/.../Maynard *Neuron* CHOSEN as second pick**, PubMed abstract + author-list used (ScienceDirect paywall not bypassed given abstract self-contained + Battle COI is the operational signal); Kleinman Lieber BDNF-DT + BDNF-AS-DT *Mol Psychiatry* CONSIDERED but paper's own conclusion is validation-limited (DROPPED); Walker NIA-led AD PWAS LILRB1 + SIRPA *Sci Transl Med* CONSIDERED but not JHU-led (JHU Bloomberg co-authors only; NIA + Alkahest corresponding — DROPPED); Cell coral photosymbiosis (Berkeley-corresponding — DROPPED); Cell 3DEEP hair follicle organogenesis (JHU BME + Dermatology; interesting platform + patent application but skin/hair biology outside core thesis — DROPPED); UMB Cell Death Dis tryptophan-branching AML (no clean spinout thesis — DROPPED); Nature astrophysics white-dwarf planet (JHU planetary science institutional match — DROPPED). Institutional press: JHTV portfolio-companies page sweep confirmed Aletira Therapeutics listed; **no Mao lab mRNA-LNP company visible; CellCipher not visible** — both are the JHTV routing prompt. Lieber Institute news landing page returns cookies consent shell via r.jina.ai (no in-window content surfaceable — same as prior day). Sources that failed transiently: initial WebFetch on incorrect DOI (10.1038/s41586-026-09765-1, PubMed-ID-derived guess) returned 404; Crossref title query returned correct DOI 10.1038/s41586-026-10729-8; r.jina.ai fetched cleanly on the correct URL. Google Search r.jina.ai proxy returned 403 on CellCipher/Battle spinout query; WebSearch tool used as fallback. Editorial call. **One dominant lead + one portfolio-adjacent second pick.** The Nature ac4C-vs-m1Ψ head-to-head with JHTV-managed LNP patent 63/325,222 is the strongest sourcing lead surfaced this month; competitive urgency is real (36-hour-old paper, every mRNA player is calling). The Neuron NAc atlas is the natural second pick because it directly informs Blackbird's Lieber GPR52 program's target biology AND Battle's CellCipher COI extends the Revivify Innovations pattern from yesterday's radar into a two-consecutive-day pattern of stealth JHU-adjacent biotech disclosures worth naming systematically. Paper links: https://www.nature.com/articles/s41586-026-10729-8 ; https://doi.org/10.1016/j.neuron.2026.06.002

Sourcing Radar — JHU Wilmer's miR-184 +57C>T paper reveals a stealth Wilmer spinout, Revivify Innovations, sitting on an anterior-segment miRNA disease-modification thesis with an iPSC-corneal-endothelium platform readout

Wed, 01 Jul 2026 12:00:00 +0000

Sourcing Radar for Wednesday, July 1, 2026. Two picks. Lead. Huang Y, Sulewski M, Xu J, Wang Y, Chen Y, Yin J, Chen J, Zack DJ, Handa JT, Al Deiri Y, Kittleson B, Bernstein AL, Perry K, Cochella L (JHU Snyder Neuroscience + Molecular Biology & Genetics), Eghrari AO (senior, corresponding; Johns Hopkins University School of Medicine, Department of Ophthalmology, Wilmer Eye Institute), "The +57C>T substitution in microRNA-184 is associated with microphthalmia, retinal detachment, and altered ocular development," medRxiv v1 2026-06-29 (DOI 10.64898/2026.06.25.26355554). Funding: Maryland Stem Cell Research Fund, JHU Alzheimer's Disease Research Center Faculty Award, NIH R01 AG066507, Research to Prevent Blindness. Competing interest — **Y. Huang, M. Sulewski, J. Xu, and A.O. Eghrari report an ownership interest in Revivify Innovations.** Background. miR-184 is the most abundant miRNA in the corneal epithelium, lens, and retina — a canonical developmental and homeostatic ocular miRNA and one of the first miRNAs implicated in a Mendelian disease (EDICT syndrome: endothelial dystrophy + iris hypoplasia + congenital cataract + stromal thinning, autosomal dominant). Cohort. 18 members of a four-generation family; 10/18 heterozygous for +57C>T. **5/10 carriers had retinal detachment vs 0/8 non-carriers (p=0.04).** Ocular biometrics carriers vs non-carriers (age+sex adjusted): axial length -2.2 mm (p=0.02, microphthalmia), mean keratometry +9.3 D steeper (p=0.004), horizontal corneal diameter -1.6 mm (p=0.0001, microcornea), central corneal thickness -139 μm (p=0.003, thinning). Mechanism. Patient iPSC-derived corneal endothelial cells (CECs) recapitulate the disease at the cellular level — irregular cell borders + increased cell/nucleus area + widened intercellular gaps + disrupted ATP1A1 (Na+/K+ ATPase) membrane localization + reduced barrier function on TEER. Gene expression: COL4A1↓ COL4A3↓ AQP1↓ COL8A1↑. Read: a seed-region SNV alters the target repertoire of miR-184, dysregulating the ECM + pump programs that maintain the corneal endothelium as a functional pump-barrier monolayer; developmental biometric phenotype implicates miR-184 in coordinated anterior-posterior eye morphogenesis. Sourcing signal. Four co-founders — first author + a Wilmer ophthalmic surgeon collaborator + a Wilmer scientist + the corresponding senior author — hold ownership in **Revivify Innovations**. AAO Financial Disclosure database independently confirms Eghrari's **PS (patent/stock)** interest in Revivify (alongside PS interests in LuckyVision + ThermOptik + GlowVD, plus consulting to Design Therapeutics + Experlo Health — Eghrari is a serial ophthalmic entrepreneur). **Revivify Innovations does not appear on the JHTV public portfolio list (A-R sweep 2026-07-01).** Three interpretations: (a) IP licensed outside JHTV / personal patent path, (b) newer than JHTV portfolio-page refresh, (c) stealth early-stage LLC. Any interpretation is a Blackbird diligence prompt. Likely thesis. miR-184-directed anterior-segment therapeutics (miRNA replacement gene therapy for LoF miRNAopathies, allele-selective knockdown for AD disease, or miR-184-pathway small molecules for corneal endothelial homeostasis) with an iPSC-CEC screening platform as shared infrastructure supporting multiple indication programs. Three angles. (1) EDICT itself is ultra-rare (single named family) but miR-184 mutation catalog extends into keratoconus + other anterior-segment developmental disorders + the retinal-detachment finding opens a new anterior-to-posterior axis. Rare disease with expansion optionality. (2) iPSC-CEC platform readouts (ATP1A1, TEER, ECM gene expression) map directly onto Fuchs endothelial corneal dystrophy screening — ~1M+ US patients, drives most corneal transplants; Aurion Biotech CLS-001 (first US corneal cell therapy) just got FDA priority review. iPSC-CEC screening is where the next generation of corneal endothelial pharmacology gets validated. (3) **Delivery adjacency to Aletira Therapeutics (Blackbird portfolio) — SELEXON is the natural cell-type-selective AAV partner for anterior-segment miRNA-replacement gene therapy.** Aletira's lead indication signal is hereditary hearing loss (inner-ear cell-type targeting); the underlying platform applies directly to corneal endothelial or lens epithelial cell targeting. Two-portfolio-company handoff worth mapping. Actions. Esther + Hemaka: courtesy JHTV route to Eghrari lab to establish Revivify Innovations state — IP posture (JHTV vs personal patent), financing stage, whether shopping seed or spoken for; this is the immediate this-week action. Geoff Lynn (Aletira): courtesy touch to Eghrari lab regardless of the equity outcome — SELEXON as delivery infrastructure for anterior-segment gene therapy is a natural partnering conversation. Avi: scope which specific miR-184 therapeutic angle Revivify is pursuing (replacement gene therapy, allele-selective ASO knockdown, or small-molecule pathway modulation) and where Blackbird chemistry could plug in. Second pick. Lateef A, Wen J, ..., Betenbaugh M (senior, corresponding; Johns Hopkins University Department of Chemical & Biomolecular Engineering), "Glycoform engineering of a mammalian platform to sculpt a humanized recombinant bioscavenger," *Cell Systems* 2026-06-30 (DOI 10.1016/j.cels.2026.101655, PMID 42379170). Compared HEK293, CHO, HepG2 for producing recombinant human butyrylcholinesterase (rHuBChE) — a stoichiometric bioscavenger for organophosphate nerve agents + pesticides. HEK293 was the best baseline; extensive glycoengineering — knocking out fucosylation, enhancing sialylation, reducing glycan branching — across >10 glycosylation-related genes achieved a near-native plasma-BChE glycoprofile. Why it matters. Native plasma-BChE has an unusually high sialic-acid + tetraantennary glycan burden critical for the >70-day plasma half-life that makes it useful as a bioscavenger dose; recombinant efforts in wild-type HEK or CHO have failed clinically because the glycan mismatch collapses half-life to hours and drives immunogenicity. Field stuck on plasma-derived material for 20+ years at limited production capacity. Blackbird angle — platform, not asset. The BChE product itself is a DoD / BARDA / BioShield SRF market — lumpy, government-driven, not a VC thesis fit. **The platform is the asset.** A HEK293 cell line with 10+ orthogonal glycoengineering edits hitting a matched native plasma glycoprofile is a general biosimilar + high-value-glycoprotein production platform. Applications: enzyme replacement therapies (Fabry α-Gal-A, Gaucher β-glucocerebrosidase, Pompe α-glucosidase — glycan-directed cellular uptake + half-life driven PK), glycan-engineered antibodies (afucosylated / bisected for enhanced ADCC), coagulation factors + fusion proteins where glycan-driven PK is the differentiator. Diligence. Betenbaugh has been at JHU 30+ years; cell-line engineering IP has historically been consumed inside industrial collaborations with AstraZeneca + MilliporeSigma. Question is whether this paper is a platform-out spin candidate or the platform is already spoken for. Avi + Hemaka to JHTV ChemBE portfolio manager to scope commercial availability. If open, exactly the kind of platform Blackbird could underwrite jointly with a downstream strategic — strategic pays for a specific glycoprotein target, Blackbird underwrites platform incubation, program funds the spin. Recommended actions consolidated. (1) Esther + Hemaka to JHTV / Wilmer on Revivify Innovations state. (2) Geoff Lynn courtesy touch to Eghrari lab on SELEXON as anterior-segment AAV delivery infrastructure regardless of equity outcome. (3) Avi + Hemaka to JHTV ChemBE portfolio manager on Betenbaugh mammalian glycoform engineering platform commercial availability + platform-out spin openness. Candidate funnel — 2026-06-28 to 2026-07-01. bioRxiv + medRxiv (api.biorxiv.org details API 2026-06-28 to 2026-07-01, paginated; 509 unique bioRxiv + 170 medRxiv records; JHU/Lieber/UMB corresponding-institution filter): 12 bioRxiv + 9 medRxiv hits. Substantive shortlist — Eghrari-Cochella miR-184 EDICT + iPSC-CEC mechanism (medRxiv 2026-06-29, JHU Wilmer + Snyder Neuroscience) **CHOSEN as lead**, full text retrieved via r.jina.ai .full route on second attempt (first attempt 2026-06-30 returned 253 chars — jina had not yet rendered the body; medRxiv rendered overnight per PIPELINE.md abstract-only-too-new pattern); NPTX2 cognitive resilience v3 (JHU Snyder, bioRxiv 2026-06-29) DROPPED as core science is from October 2025 v1; instrumental-motor transfer v1+v2 (JHU SOM Neuro, bioRxiv 2026-06-28+29) DROPPED as basic neuroscience; antifungal use in septic shock v1 (JHU SOM ICU, medRxiv 2026-06-30) DROPPED as clinical utilization epi; statins in SLE CV events v1 (JHU Cardiology, medRxiv 2026-06-30) DROPPED as clinical epi; genetic counseling videos v1 + countdown paradox MCI biomarker clock v1 (JHU, medRxiv 2026-06-29) DROPPED (previously; educational + epidemiological). PubMed (JHU/UMB/Lieber affiliation entry-date 2026-06-28 to 2026-07-01): 126 unique records; therapeutic-flagged after dropping reviews + public-health surveillance — Betenbaugh glycoform-engineered rHuBChE (*Cell Systems* 2026-06-30, JHU ChemBE) **CHOSEN as second pick**, full text behind Cell Press paywall + ScienceDirect Cloudflare CAPTCHA (PubMed abstract used with detailed glycan modification detail); Berger flex-fuel replisomes (*Nat Commun*, JHU Biophysics) DROPPED as basic mechanism; Jaffee apCAF / CCL22 in PDAC (*J Immunol*, JHU Convergence Institute) DROPPED — CCL22 has existing pharmacology (CCR4 antagonists incl mogamulizumab), no IP angle as standalone sourcing lead; Azad atezolizumab+varlilumab±cobimetinib in BTC (*CCR*, JHU SKCCC) DROPPED as negative Ph2; Solomon HCV support trial (*J Hepatol*, JHU Bloomberg Epi) DROPPED as public health; Handa Wilmer CAR-T ocular case report (*Front Immunol*) DROPPED as case report; JHU Neurosurgery tDCS depression protocol DROPPED as feasibility protocol. High-tier journal sweep (Cell/Nature/Science/Nat Med/Nat Biotech/Sci Transl Med/Cell Rep Med/JCI/Sci Adv/Nat Commun/Cell Metab/Cancer Cell/Cell Stem Cell/Immunity/Neuron/Nat Neurosci/Nat Struct Mol Biol/Nat Cell Biol/Sci Immunol/NEJM/JAMA/Lancet/Nat Chem Biol/Cell Chem Biol/Nat Genet, JHU/UMB/Lieber pdat 2026-06-28 to 2026-07-01): 3 hits — Berger *Nat Commun* replisomes (dropped as above); BridgeBio Ph3 infigratinib in achondroplasia *NEJM* (BridgeBio-corresponding, not JHU-led — dropped); UVA-led glioblastoma miRNA *JCI* (not JHU-led — dropped). Institutional press (JHTV, hub.jhu.edu, hopkinsmedicine.org, libd.org, ventures.jhu.edu, umventures.org via r.jina.ai proxy): JHTV in-window items are DecisionRx × Bloomberg School collaboration (previously covered/dropped), Bayh-Dole Innovator Award recap on Dannals + Pomper Pylarify (published 2026-06-22, retrospective on 2021-approved product), Pava Center + Social Innovation Lab spring 2026 showcase (student-stage). Hub.jhu.edu in-window: JHU Press awards + Milestones July 2026 (housekeeping). Hopkins Medicine newsroom: patient story + Fourth-of-July burn safety (non-research). Lieber Institute news: cookies consent shell via r.jina.ai (no surfaceable content). UM Ventures news: pre-window (GEn1E Duchenne + Secretome $30M Series A both 2026-06-09 and earlier). Sources that failed transiently: Google + Bing r.jina.ai proxies polluted on "Revivify Innovations" query (campground + engine-manual pages — anti-scraping noise, not source failures; AAO Financial Disclosure database independently confirmed Eghrari + Revivify + PS relationship); doi.org r.jina.ai proxy SecurityCompromiseError HTTP 451 due to Cloudflare abuse rate-limit (PubMed abstract used for BChE); ScienceDirect r.jina.ai proxy Cloudflare CAPTCHA on Cell Systems article (PubMed abstract used); www.medrxiv.org direct fetch on miR-184 EDICT DOI required r.jina.ai proxy — first attempt 2026-06-30 returned 253 chars (jina had not rendered body), second attempt 2026-07-01 returned full 5.3k-char body. Editorial call. Two-pick day. The miR-184 EDICT + Revivify Innovations story is the strongest partner-institution-anchored translational item in window — the mechanism paper is peer-reviewable quality, but the operational lead is the *company disclosure* that promotes it from a rare-disease mechanism study to an active-company sourcing conversation with direct Aletira SELEXON delivery adjacency. The Betenbaugh Cell Systems paper is added as a brief second pick because JHU-led platform technology with credible spin-out openness is exactly the second-tier item the radar should surface. Neither displaces the other; the two are complementary — asset-adjacent company signal + platform-out spin candidate. Paper links: https://www.medrxiv.org/content/10.64898/2026.06.25.26355554v1 ; https://doi.org/10.1016/j.cels.2026.101655

Sourcing Radar — Johns Hopkins angiogenin / mitoribosome biogenesis paper opens a bone-and-muscle pharmacology lane for an ALS-anchored protein, with two clean indications and a JHU Orthopaedic translational scaffold

Tue, 30 Jun 2026 12:00:00 +0000

Sourcing Radar for Tuesday, June 30, 2026. Two picks plus a portfolio note. Portfolio note. JHTV posted today a detailed feature on **1104health** explicitly identifying the company as a Blackbird BioVentures portfolio investment — JHTV-licensed know-how, founder Rose Wang (engineer + serial entrepreneur, founded after her husband died of cancer in 2019), JHU Community Clinical Research Network + melanoma group (Lipson + Warrier) collaboration, pre-launch as of June 2026 with broader rollout in spring 2026. Pivoted from a patient-facing model to a clinician-focused marketplace embedding clinical-trial options in community-oncology workflow with pharma-funded subsidies for trial-coordination work. First detailed public confirmation of the trial-enrollment platform that was previously unnamed in our portfolio reference. Saved for Sunday Portfolio Watch (2026-07-05); portfolio memory updated. Lead pick. Shen K, Zeng Y, Wang J, Song K, Kumar S, Gao P, Hu G-F (Tufts), Cao X, Wan M (senior, corresponding; Johns Hopkins Orthopaedic Surgery), "Angiogenin mediates cell-cell fusion as a mitochondrial RNA processing enzyme," *Bone Research* 14 (2026), DOI 10.1038/s41413-026-00545-1 (PMID 42374020), online 2026-06-29. Funding: NIA R01AG068226 (Wan), R01AG072090 (Wan/Cao), P01AG066603 (Cao). Background. Angiogenin (RNase 5) is the founding member of the RNase A family — long studied as a secreted angiogenic factor, an ALS-mutation gene (Greenway/Cudkowicz 2006; multiple LoF variants — R31K, R121H, K40I, K17I), and a stress-induced cytoplasmic ribonuclease generating tiRNAs from tRNAs to suppress translation under stress. ANG-mimetic peptides have been in early development for ALS out of Guo-Fu Hu's lab at Tufts (co-author on this paper); the ALS therapeutic window is hostile and the programs have not progressed past early proof-of-concept. New mechanism. Upon initiation of cell-cell fusion, ANG translocates from nucleus to mitochondria where it acts as a mitochondrial-tRNA 3'-end processing enzyme. Transcriptome-wide PARE + 5' RACE show ANG cleaves mt-tRNA 3' ends at junctions where the tRNAs border mt-rRNAs and mt-mRNAs in mitochondrial polycistronic precursors — releasing the rRNAs/mRNAs for mitoribosome assembly and translation of respiratory complex subunits. The catalytically dead K40I ALS-mutant ANG does not rescue. Mechanism depends on the ribonucleolytic activity, not the secreted-angiogenic function. In vivo phenotypes. Ang-/- mice have elevated BMD + elevated trabecular bone volume with markedly reduced osteoclast numbers + surface area (high bone mass driven by suppressed osteoclast-precursor fusion). Ang-/- mice also have reduced gastrocnemius weight, smaller myofiber CSA, fewer myonuclei per myofiber, and a 45.98% reduction in treadmill exhaustion total work vs wild-type. Local viral wtANG injection robustly rescues BaCl2 muscle injury regeneration; K40I mutant does not. Bidirectional translational read. **Inhibit ANG** for osteoporosis (rationale: suppress osteoclast precursor fusion → reduce resorption; directionally bisphosphonate / denosumab-like with distinct mechanism). **Activate ANG / ANG mimetics** for sarcopenia + skeletal muscle regeneration disorders (rationale: enhance myoblast fusion during satellite-cell-mediated regeneration). IP angle is method-of-use rather than CoM — molecules already exist (recombinant ANG, ANG-mimetic peptides from Hu's lab, small molecules from earlier ALS screens) — the new IP is the indication: ANG modulators for osteoporosis and for sarcopenia / muscle regeneration. Markets. Anti-resorptive: bisphosphonates (generic, $4B+ legacy), denosumab (Amgen Prolia / Xgeva, $5B+ peak), romosozumab (Amgen Evenity); postmenopausal osteoporosis remains 10M+ US patients; denosumab rebound resorption + atypical femur fractures leave room for differentiated mechanism. Sarcopenia: no approved drug; recent activity around myostatin/activin pathway (apitegromab — Scholar Rock for SMA; bimagrumab — Lilly for GLP-1 muscle loss; trevogrumab — Regeneron). The GLP-1-induced lean-mass-loss problem has created a renewed therapeutic window for muscle-anabolic agents. ANG-mediated myoblast-fusion biology is mechanistically distinct from the myostatin axis and could be combined. JHU Orthopaedic Surgery as a translational department is one of the strongest bone-and-muscle programs nationally — Cao + Wan + the NIA R01+P01 aging-program scaffolding. Diligence questions. Hemaka — courtesy-route JHTV on the Wan-Cao group: are bone+muscle method-of-use claims on ANG modulators being filed or open territory? Parallel ping to Hu's group at Tufts about existing ANG peptide chemistry available for re-positioning licensing into bone/muscle indications. Second pick. Kumari A, Nguyen DM, Disilvestre D, Dirda NDA, Kethanapalli SH, Kao JPY, Garg V (corresponding; UMB SOM Pharmacology and Physiology), "Mechanistic basis of EMRE's essential role in the regulation of mitochondrial calcium uniporter complex," bioRxiv v1 2026-06-29, DOI 10.64898/2026.06.25.733848. Funding: NIH R01GM145806, AHA 24POST1241582, NIAMS T32 AR007592, Maryland Stem Cell Research Fund MSCRF-96651. New high-sensitivity assay quantifying MCU complex (MCUcx) function in intact mitochondria, paired with mechanistic results establishing two distinct roles for EMRE: (i) matrix-Ca2+-dependent inhibition of MCU through a juxtamembrane site, mechanistically distinct from MICU1-mediated low-cytosolic-Ca2+ inhibition; (ii) EMRE actively promotes ion permeation through MCU, elevating its role from structural scaffold to active throughput determinant. Stated translational frame from the authors: "highlight the utility of our assay for probing MCUcx biophysical mechanisms and enabling the discovery of uniporter modulators." Why it matters. MCU pharmacology has been a stalled space — known modulators (Ru360, mitoxantrone, MCU-i4 derivatives) lack drug-like properties; the field has been blocked by lack of a high-throughput intact-mitochondrion functional assay. Cardiac IR-injury rationale (Ca2+ overload triggers PTP opening + cardiomyocyte death) and neurodegeneration rationale (mitochondrial Ca2+ dysregulation in PD, AD) are old hypotheses awaiting tool compounds. Blackbird structural fit. UM Ventures × Blackbird co-investment agreement is the natural framing; platform-stage rather than asset-stage; exactly the kind of MCU-modulator program UMB + Blackbird could underwrite jointly. Diligence question for Esther — confirm with UM Ventures (Robilotto) whether Garg lab is tracked as potential spin-out target, and whether Garg is open to a Blackbird translational-grant pilot to deliver tool compounds + cardiac IR readout. Recommended actions consolidated. (1) Hemaka to JHTV on Wan-Cao bone/muscle ANG IP posture + Hu/Tufts ANG chemistry re-positioning options. (2) Esther to UM Ventures on Garg EMRE MCU spin-out posture + receptiveness to Blackbird-grant pilot. (3) Avi to internal scope on whether either ANG or MCU pharmacology has Blackbird-incubatable chemistry angles separate from legacy assets. Candidate funnel — 2026-06-27 to 2026-06-30. bioRxiv + medRxiv (api.biorxiv.org details API, paginated, 2026-06-27 to 2026-06-30, 540 unique preprint/version records, JHU/Lieber/UMB filter on corresponding institution + authors + abstract): 8 hits. Substantive shortlist — Wan-Cao ANG paper (CHOSEN as lead — Bone Research peer-reviewed, JHU Orthopaedic-corresponding); Garg EMRE/MCU UMB Pharmacology v1 2026-06-29 (CHOSEN as second pick — UM Ventures × Blackbird structural fit); Worley/Lao NPTX2 multi-omics v3 2026-06-29 (DROPPED — core science is from v1 October 2025; correlative integration, no new asset); Eghrari/Cochella microRNA-184 microphthalmia JHU Wilmer (DROPPED — full text unavailable, jina returned 253 chars on a <24h preprint); Al-Fawakhiri instrumental-motor transfer v2 JHU SOM (DROPPED — basic neuroscience, no drug angle); Farber LipoGlo v5 (DROPPED — already covered Friday 2026-06-26); Brown JHU genetic-counseling videos medRxiv (DROPPED — education tool); Zhu/Pettigrew countdown paradox MCI biomarker clock medRxiv (DROPPED — epidemiology). PubMed (JHU/UMB/Lieber affiliation entry-date 2026-06-27 to 2026-06-30): 144 records, therapeutic-flagged after dropping spurious + already-covered — Shen/Wan ANG paper (PMID 42374020, *Bone Res*, JHU Orthopaedic) confirms the bioRxiv lead in a peer-reviewed venue; other JHU/UMB-led therapeutic items dropped — Yakovlev/Medved N-cadherin × fibrin angiogenesis Thromb Haemost UMB Biochemistry (basic mechanism, no immediate translational lane), Ansari/Nyquist Treg acute brain injury Mol Neurobiol JHU NCC (review article), Rashid/Berger replisomes Nat Commun JHU Biophysics (basic mechanism), Li/Nakazawa Th17 anti-PD1 signature Cancer Immunol Res JHU SKCCC (clinical biomarker, no IP angle), Chen/Cheng GeoPep peptide binding ML J Chem Inf Model JHU BME + OSU (ML framework, not an asset), Erel-Akbaba BATUS ultrasound gene therapy Adv Sci (UConn-led, JHU Theodore + Doloff contributor not lead), Islam/Segars Src+β-catenin uterine fibroid JHU OBGYN (covered Sunday 2026-06-28), Rehg/Miranda-Barrientos Lieber mPFC-LC Neuropsychopharmacology (covered Sunday 2026-06-28). High-tier journal sweep (Cell/Nature/Science/Nat Med/Nat Biotech/Sci Transl Med/Cell Rep Med/J Clin Invest/Sci Adv/Nat Commun/Cell Metab/Cell Stem Cell/Immunity/Cancer Cell/Neuron, JHU+UMB+Lieber affiliation, pdat 2026-06-27 to 2026-06-30): 7 hits, none displaceable as JHU/UMB/Lieber-led therapeutic sourcing leads (retraction Indiana-led HER2; JHU Berger replisomes biophysics; Pittsburgh-led bacteriophage; MD Anderson-led BCL2 BRAFi-MEKi melanoma; Minnesota-led MMR-deficient prostate; NIA-led GDF15 dementia; JHU ChemE eDAC non-biomedical). Institutional press (JHTV, hub.jhu.edu, hopkinsmedicine.org, libd.org, ventures.jhu.edu, umventures.org via r.jina.ai proxy): hub.jhu.edu in-window items are housekeeping (Milestones July 2026; payroll reminders). JHTV today (2026-06-30) features three startup stories — 1104health (Blackbird portfolio — saved for Sunday Portfolio Watch + portfolio memory updated); SmartTIVA closed-loop anesthesia delivery (JHU master's-in-engineering founder Juan Fernandez, pre-launch, medtech outside Blackbird thesis — dropped); JHU Bloomberg School × DecisionRx pharmacogenomics-led medication optimization (population-health analytics + value-based care collaboration — not a therapeutic sourcing lead). Sources that failed transiently: nature.com direct article fetch (idp.nature.com auth wall — used r.jina.ai proxy on the *Bone Research* article URL to confirm in vivo phenotypes); bioRxiv direct fetch on 10.64898/2026.06.25.26355554 returned 253 chars (preprint <24h old, jina hadn't rendered the body); PubMed eFetch HTTP 429 once on high-tier sweep (no information lost). Editorial call. Two-pick day. The Wan-Cao angiogenin paper is the strongest JHU-led translational item in window — peer-reviewed venue, clean in vivo bone+muscle phenotypes with viral rescue, two large indication markets, method-of-use IP path faster than de novo discovery. The Garg EMRE/MCU paper is added as a brief second pick — UMB-led platform technology with explicit UM Ventures × Blackbird structural fit. 1104health JHTV portfolio piece is correctly held for Sunday Portfolio Watch rather than miscategorized into today's Sourcing Radar. Paper links: https://www.nature.com/articles/s41413-026-00545-1 ; https://www.biorxiv.org/content/10.64898/2026.06.25.733848v1

Sourcing Radar — Johns Hopkins PAH metabolome: a fatty-acid-oxidation axis predicts adverse RV-PV physiology, and a transpulmonary 5α-reduced androgen gradient is protective — two pharmacological lanes in the PAH market that sotatercept did not close

Mon, 29 Jun 2026 12:00:00 +0000

Sourcing Radar for Monday, June 29, 2026. One pick. Clinton I, PVDOMICS Study Group, Coursen J, Rosen D, Suresh K, Balasubramanian A, Kolb TM, Damico RL, Mathai SC, Hsu S, Mukherjee M, Finet JE, Grunig G, Barnard J, Hemnes AR, Leopold JA, Horn EM, Rosenzweig EB, Rischard F, Frantz RP, Erzurum S, King W, Beck G, Hill NS, Hassoun P, Simpson CE (senior, corresponding; Johns Hopkins University School of Medicine, Division of Pulmonary and Critical Care Medicine), "Metabolomic Network Analysis Reveals Reorganization of Lipid and Steroid Programs Linked to Right Ventricular-Pulmonary Vascular Function in Pulmonary Hypertension," bioRxiv v1 posted 2026-06-22 (DOI 10.64898/2026.06.16.732773). Funding NIH R03 HL176624 + R01 HL175011 (Simpson), R01 HL172830 (Hsu). Discovery cohort: PVDOMICS multicenter consortium, n=412 PAH. Replication cohort: CALIPSO, single-center Johns Hopkins Hospital, n=89. Market framing. PAH is ~50-60k US patients. Standard of care: ERAs, PDE5 inhibitors, prostacyclin pathway, sGC stimulators, plus sotatercept (Merck Winrevair, approved 3/2024, peak forecast $3-5B+). 5-yr mortality still ~40-50%. RV dysfunction is what kills these patients; no approved therapy directly addresses RV metabolism. That is the gap this paper points at. What the paper does. WGCNA applied to untargeted Metabolon HD4 plasma metabolomics, 412 PAH patients, replicated on the 89-patient JHU CALIPSO biobank. 16 distinct, biologically coherent metabolite modules; core architecture and conserved hub metabolites recovered in replication. Two clean signals. (1) Tripartite fatty-acid-oxidation axis: brown module (substrate pool — long-chain unsaturated FAs, odd-chain FAs, medium-chain FAs); blue module (long-chain acylcarnitines + dicarboxylic FAs — mitochondrial β-oxidation + peroxisomal ω-oxidation products); green module (acyl-glycine/acyl-glutamine conjugates + hydroxyacylcarnitines — disposal products). Sequential FA handling from uptake → oxidation → disposal. Blue + green strongly associated with adverse hemodynamics: per IQR increase in blue eigenmetabolite, PVR +1.69 WU [1.12–2.26] and mPAP +6.24 mmHg [4.42–8.06] and RVEF -2.05%/IQR; per IQR green, PVR +1.26 WU and mPAP +4.51 mmHg and RVEF -2.61%/IQR. Persisted with fluid and vasodilator challenge. Substrate pool itself not strongly associated. Translational reading: increased FA flux with incomplete oxidation and alternative disposal is a metabolic signature of the struggling pressure-overloaded RV. (2) Steroid metabolism axis splits into two related modules. Yellow (precursor — contains DHEA-S, the canonical PAH sex-hormone biomarker): *no* association with RV-PV indices. Cyan (5α-reduced androgen derivatives — downstream conversion products): clearly protective, lower right-sided pressures and improved RV-PA coupling. Transpulmonary gradient analysis (multi-site sampling across pulmonary circulation + peripheral vein): yellow precursor module shows transpulmonary uptake, cyan reduced-androgen module shows release. The pulmonary vasculature is doing the conversion. This explains the EDIPHY trial failure cleanly — randomized placebo-controlled crossover DHEA supplementation in PAH showed no RV longitudinal strain improvement despite robust DHEA-S increases, because supplementing the precursor doesn't help if the conversion is rate-limiting. Therapeutic target is the conversion machinery (5α-reductase or downstream pathway enzymes in pulmonary endothelium), not the circulating precursor pool. Novel, actionable pharmacology target lane that nobody is currently developing on for PAH. Three Blackbird reads. (1) Sourcing lead, not portfolio. Blackbird does not currently have a PAH program; PAH is a large underserved cardiovascular market with one new disease-modifier in fifteen years. JHU has structural advantage — co-leads PVDOMICS, owns CALIPSO single-center biobank, has the RHC + cardiac MRI + Metabolon + clinical phenotyping integration in-house. The kind of clinical-translational scaffolding that produces investor-ready preclinical packages on a 6-18 month horizon. Worth Esther + Hemaka adding the Simpson-Hassoun-Mathai group to the partner-institution radar even absent immediate action. (2) Two pharmacological lanes flagged in the data. Lane one — FAO pathway modulators in PAH with the FAO module as a clinical-trial enrichment biomarker. Trimetazidine (Servier, EU/Asia angina approval) and ranolazine (Gilead Ranexa, FDA angina approval) are existing partial FAO inhibitors with cardiac history but neither has been run in a properly biomarker-enriched PAH cohort. Etomoxir failed historically for hepatotoxicity, leaving novel chemistry on selective CPT-1 inhibitors or 3-ketoacyl-CoA thiolase inhibition as open territory. Lane two — 5α-reductase pathway pharmacology in the pulmonary endothelium, *opposite* directionality from the prostate finasteride-dutasteride class. Novel pharmacology, no clean precedent, the question is whether facilitating or augmenting pulmonary-endothelial 5α-reduction is tractable as a small-molecule or biologic target. Both lanes pre-target-validation today and would benefit from an in-vivo PAH rodent program plus a PK/PD readout in a CALIPSO-style biobank — exactly the work a Blackbird translational grant funds. (3) Diligence question for Hemaka. What is the JHTV relationship with the Simpson-Hassoun group at the moment — shopping a target hypothesis to a strategic, queuing a follow-on paper with in vivo readouts, or at the open-territory stage where a Blackbird-grant-supported in-vivo + lead-discovery package would be the right intervention? Determines whether this is a near-term sourcing conversation or a one-year-out watch-item. Worth a courtesy JHTV-routed outreach this week. Recommended actions. (1) Esther + Hemaka add the Hopkins PAH translational group (Simpson, Hassoun, Mathai anchor) to the partner-institution sourcing radar. (2) Hemaka courtesy-routes JHTV on the stage and posture of the Simpson program with a specific question about an in vivo target-validation work package on the FAO or 5α-reductase lanes through a Blackbird translational grant. (3) Avi internal-scopes whether either pharmacology lane has any composition-of-matter angle a Blackbird-incubated chemistry effort could anchor, separate from legacy FAO partial-inhibitor or finasteride-class molecules. Candidate funnel — 2026-06-26 to 2026-06-29. PubMed (Johns Hopkins/Lieber/UMB affiliation entry-date 2026-06-26 to 2026-06-29; 107 records; 67 metadata-fetched, therapeutic/target-discovery/translational filter): mostly clinical, epidemiology, surveys, reviews. Substantive shortlist after dropping spurious matches: Sci Adv Plasma GDF15 dementia (Walker NIA-led, JHU contributors only — DROPPED, not JHU-led); Cell Rep Med MMR-deficient prostate ICR (Antonarakis Minnesota-led, single case, JHU SKCCC contributors only — DROPPED); Cancer Cell intratumor phage-fibroblast-B cell circuit (Zeng Weill Cornell-led, JHU Egeblad co-author — DROPPED); Acta Neuropathol Commun NEK1 missense ALS (Brenner Ulm-led, JHU Aly peripheral — DROPPED); Cell plasma lung tumor promotion (Swanton Crick-led, JHU Bloomberg co-authors only — DROPPED); Eur J Med Chem MOR-D3R etonitazene dual ligands (Newman NIDA-IRP-led, JHU Drug Discovery Rais peripheral — DROPPED); MDPI Pharmaceuticals MEIS2 breast cancer (Kocabaş Istanbul-led, JHU All Children's FL peripheral — DROPPED). bioRxiv (api.biorxiv.org details API, 2026-06-22 to 2026-06-29; 1014 unique DOIs; corresponding-institution filter Hopkins/Lieber/UMB): 11 hits. Substantive shortlist: Simpson PAH metabolome (10.64898/2026.06.16.732773, JHU SOM, physiology, v1 2026-06-22) — **CHOSEN as lead**, full text via r.jina.ai .full route; Casadevall AmpB resistance / Lomentospora prolificans (10.64898/2026.06.25.734450, JHU SOPH, microbiology, v1 2026-06-25) — interesting mechanism (soluble >100 kDa cell-wall component inactivates AMB) but market too narrow (~0.5 cases/million/yr) for typical biotech — DROPPED; Farber LipoGlo ApoB phenotypic screen (10.1101/2024.11.14.623618, JHU, physiology, v5 2026-06-27) — minor revision of v4 covered Friday 2026-06-26 — DROPPED; Simonetti agQVOA HIV antigen-stim latency reversal (10.64898/2026.06.11.731680, JHU SOM, immunology, v2 2026-06-23) — same paper covered Tuesday 2026-06-24 — DROPPED; Pergola D2 reward imaging (10.64898/2026.05.14.724267, Lieber, neuroscience, v3 2026-06-24) — adjacent to sign-tracker biomarker covered Friday 2026-06-26 — DROPPED; Zimin EviAnn (JHU, bioinformatics tool) — DROPPED; Al-Fawakhiri instrumental-motor learning (JHU SOM, neuroscience, basic) — DROPPED; Foster turtle shell muscle loss (JHU SOM, evolutionary), Carroll snake venom (UMD College Park *not* UMB), Danilova replicateFest (JHU SOM, immunology software, covered Sat 2026-06-27), Yovanno directional info flow (JHU SOM, biophysics methodology, covered Wed 2026-06-24) — all DROPPED. High-tier journal sweep (Cell/Nature/Science/Nat Med/Nat Biotech/Sci Transl Med/Cell Rep Med/JCI, JHU affiliation pdat 2026-06-22 to 2026-06-29): 3 hits, all JHU contributor not lead, none displaceable as JHU sourcing leads. Institutional press (hub.jhu.edu, hopkinsmedicine.org, libd.org, ventures.jhu.edu, umventures.org): Cloudflare 403s direct, r.jina.ai proxy returns CAPTCHA/cookie shells — no surfaceable JHU/Lieber/UMB institutional news in window. UMB Grid Pitch 2026 surfaced via web search but content is student-stage ventures (iMPILO Therapeutics, Benign) too early for Blackbird sourcing. Sources that failed transiently: bioRxiv search HTML (Cloudflare CAPTCHA on direct fetch and on r.jina.ai proxy — api.biorxiv.org details endpoint used as bypass per PIPELINE.md). Europe PMC affiliation-filtered preprint search returns no in-window records (indexing lag of weeks-to-months on preprints — confirmed by sort=FIRST_PDATE+desc returning 2024-06 records as most-recent JHU preprint). Editorial call: Simpson PAH paper is the single highest-quality Hopkins-led translational item in window, with two named pharmacological lanes and Hopkins consortium-leadership structural advantage. Single-pick day is consistent with PIPELINE.md "thin day is a short episode" guidance — Casadevall Lomentospora and Farber LipoGlo v5 do not clear the translational-investor bar as standalone sourcing leads on a non-portfolio-touching day. Paper link: https://www.biorxiv.org/content/10.64898/2026.06.16.732773v1

Sourcing Radar — Lieber dissects an mPFC→LC top-down attention circuit, with explicit handoff to neuropsychiatric drug development

Sun, 28 Jun 2026 12:00:00 +0000

Sourcing Radar for Sunday, June 28, 2026. Lead. Rehg JJ, Olivares DE, Li Y, García R, Martinowich K, Carr GV, Miranda-Barrientos J (corresponding; Lieber Institute for Brain Development + JHU Department of Psychiatry and Behavioral Sciences), "Top-down control of sustained attention by medial prefrontal cortex-locus coeruleus (mPFC-LC) projection neurons during the rodent continuous performance test (rCPT)," Neuropsychopharmacology, posted 2026-06-26 (DOI 10.1038/s41386-026-02452-9, PubMed PMID 42362777). All senior authors dual-appointed Lieber + Johns Hopkins SoM (Neuroscience / Physiology-Pharmacology-Therapeutics / Psychiatry); Greg Carr was previously head of medicinal chemistry / in vivo pharmacology at the Lieber preclinical platform. The science. Selective chemogenetic activation of mPFC-to-LC projector neurons in the rodent continuous performance test (rCPT) — a translational sustained-attention paradigm cross-mapped to the human CPT used in ADHD/SCZ/AD attention batteries — improves discrimination (d' increase), the cognitive-precision axis. Non-selective activation of all mPFC neurons improves performance via increased responsivity (more hits, more false alarms, no d' change), engaging nucleus accumbens rather than LC. Two prefrontal subpopulations, two functionally dissociated attention domains, non-overlapping downstream targets. Microcircuit: mPFC-LC projectors recruit LC-NE neurons plus glutamatergic and GABAergic peri-LC neurons; net effect is a closed-loop increase in NE tone within the mPFC itself. Therapeutic framing (verbatim from paper abstract): "providing a circuit-level framework for understanding the mechanisms of sustained attention and for developing targeted therapies for attentional deficits across neuropsychiatric disorders." Cross-indication: ADHD, schizophrenia cognitive symptoms, age-related attention decline. Blackbird angle — three reads. (1) Strategic frame for the GPR52 NewCo with Lieber and Third Rock. The GPR52 thesis hits all three SCZ symptom domains — positive, negative, cognitive — without dopamine. Cognitive symptoms anchor in attention/working-memory. This paper opens a complementary non-dopaminergic anatomically-targetable circuit for the cognitive piece, with a closed-loop NE-amplification mechanism that does not require the blanket noradrenergic exposure limiting Strattera (Otsuka atomoxetine) or Supernus SPN-812 viloxazine. Conversation for Esther + Hemaka with Lieber: is the Carr-Martinowich LC pipeline running independently of the GPR52 NewCo, or could it become a second portfolio asset under the same Lieber-Blackbird partnership umbrella? (2) Platform read. Third Lieber infrastructure asset in three weeks — Pergola trans-TWAS INGENE/MODULE target-discovery engine (06-25), Pergola sign-tracker/goal-tracker fMRI biomarker (06-26), and now this LC-circuit story (06-28). The case for a platform-layer Lieber-Blackbird relationship — multiple single-asset spinouts riding on a shared Lieber neuropsychiatric-platform stack — is substantially stronger than a month ago. Hemaka to Tom Hyde at Lieber. (3) Diligence question. The chemogenetic intervention is a research tool, not a drug; translational version is a small-molecule modulator or circuit-specific gene-therapy tool that selectively activates the mPFC-LC projector population without globally elevating NE tone. Avi to ask Carr directly whether the lead-discovery work behind this circuit handoff is already underway in-house at Lieber drug discovery, or unstarted and open territory. Second item. Islam MS, Macri VI, Martin L, Shen C, Mejias J, El Sayed S, Michel R, Post C, Singh B, Brennan JT, Fennell A, Phillip JM, Borahay MA, Elisseeff JH, Segars JH (senior, corresponding; Division of Reproductive Sciences and Women's Health Research, Department of Gynecology and Obstetrics, Johns Hopkins Medicine + Translational Tissue Engineering Center, Johns Hopkins), "Src and β-catenin control a senescence-fibrotic phenotype in uterine fibroid cells," Scientific Reports, posted 2026-06-26 (DOI 10.1038/s41598-026-57742-5, PubMed PMID 42362678). Market. Uterine fibroids hit ~70% of women by menopause, drive most of the ~600k/yr US hysterectomies. Only FDA-approved Rx are three GnRH-antagonist combos — AbbVie's Oriahnn (elagolix), Pfizer-Sumitomo's Myfembree (relugolix/E2/NETA), Gedeon Richter's Yselty (linzagolix) — all symptom-control hormonal Rx, finite duration use due to bone-density warnings. No FDA-approved disease-modifying agent on the fibrosis biology. Mechanism. Substrate-stiffness sufficient to upregulate senescence (p16, p21) plus fibrotic markers (fibronectin, PAI-1, αSMA) in fibroid cells. Src kinase + β-catenin are the regulatory nodes. Dasatinib (approved BCR-ABL/Src inhibitor) attenuates senescence + fibrosis markers + shuts down β-catenin → Src is upstream. β-catenin inhibitors ICG-001 + PRI-724 (Prism BioLab / Eisai chemistry) phenocopy at senescence and fibrosis readouts, knock down CTGF and cyclin-D1. Biomechanics → senescence → fibrosis, with Src-β-catenin the druggable node. Commercial caveat. Dasatinib + the legacy β-catenin chemistry are not composition-of-matter defensible in fibroids. Patentable angles: novel Src-selective or β-catenin-pathway chemistry tuned to uterine tissue context, or a use-patent on the senolytic + anti-fibrotic combination in fibroids. Borahay + Segars sit in the broader JHU Drug Discovery ecosystem adjacent to Slusher; the Sci Reports placement suggests the next paper — with in vivo / preclinical PK/PD readouts — is in preparation. Blackbird action. Hemaka courtesy outreach via JHTV to the Segars group: where is the program on the translational arc, is lead-discovery already underway, is the team receptive to a Blackbird-grant-supported in-vivo + IND-enabling work package. Recommended actions consolidated. (1) Esther + Hemaka to Lieber on the Carr-Martinowich LC pipeline as candidate second portfolio asset complementary to GPR52. (2) Avi to scope whether the medchem handoff on the mPFC-LC circuit is in-house at Lieber or open. (3) Hemaka to JHTV on the Borahay-Segars fibroid program — courtesy diligence on stage, in vivo plans, openness to Blackbird translational grant. Candidate funnel — 2026-06-25 to 2026-06-28. PubMed (Johns Hopkins/UMB/Lieber affiliation × therapeutic/novel-target/platform/translational/drug discovery/small molecule/inhibitor/agonist/biomarker/vaccine/gene therapy/antibody, 17 + 20 returned, 25 metadata-fetched): Lieber Rehg/Carr/Miranda-Barrientos mPFC-LC sustained attention (CHOSEN as lead); JHU Segars/Borahay/Elisseeff uterine fibroid Src/β-catenin (CHOSEN as second pick); JHU Ding (review co-author) lineage-switch acute leukemia Leukemia review (Stanford-led not JHU-led, dropped as review); MD Anderson Y N/Davies BCL2 family BRAFi-MEKi melanoma resistance Nat Comm (MD Anderson-led, JHU Vito Rebecca middle author Bloomberg School, dropped); JHU Aly NEK1 missense ALS Acta Neuropathol Commun (Ulm-led with JHU peripheral, dropped); Pharmaceuticals MEIS2 small-molecule breast cancer (JHU All Children's St. Petersburg FL author, MDPI Pharmaceuticals — peripheral, lower-tier journal, dropped); Minnesota-led MMR-deficient prostate cancer case report Cell Rep Med (single-case, JHU supporting only); GDF15 dementia Sci Adv (NIA-led, JHU contributors only); JHU Wilmer AMD drug delivery review (review-only, dropped); other JHU PubMed hits non-translational. bioRxiv (search_preprints 2026-06-25 to 2026-06-28; pharmacology/cancer/neuroscience/bioengineering/immunology, ~80 in-window): no JHU/UMB/Lieber-anchored translational items beyond what was caught by PubMed; cancer biology 0 hits in window. UMB Lieber bioRxiv: no anchored hits. Institutional press (hub.jhu.edu, hopkinsmedicine.org, libd.org, ventures.jhu.edu, umventures.org, medschool.umaryland.edu): JHTV 2026 Celebration of Innovation in Medicine + Slusher Dean's Distinguished Faculty Innovator Award — recap article dated 2026-05-05, out of 3-4-day window, dropped from Radar but partner-institution color for record. Industry press: nothing portfolio-touching that displaces these picks for Sourcing Radar (Portfolio Watch items separately covered in Sunday companion episode). Sources that failed or were proxied: hub.jhu.edu + hopkinsmedicine.org (Cloudflare 403, used r.jina.ai); www.nature.com/articles/s41386-026-02452-9 direct fetch redirected to idp.nature.com auth wall, abstract used. Already shipped this week (excluded): DCTPP1 decitabine (06-21), Wilmer crystal sorafenib AMD (06-22), epigenetic vaccine sarcoma + OBM sickle-cell (06-23), agQVOA HIV deep latency (06-24), Lieber trans-eQTL INGENE/MODULE schizophrenia targets (06-25), Lieber sign-tracker/goal-tracker D2 biomarker + JHU LipoGlo (06-26), Hopkins mKRAS-VAX MMRP CRC + replicateFest (06-27). Editorial call: Lieber mPFC-LC paper is the single best partner-institution-anchored translational item in window — Lieber-led, named drug-development handoff in the abstract itself, direct portfolio-platform read on the GPR52 NewCo. Second pick on JHU fibroid mechanism rounds out a 2-pick Sunday Radar that fits the "short Radar plus Portfolio Watch" Sunday format. Paper link: https://www.nature.com/articles/s41386-026-02452-9

Portfolio Watch — Sangamo bankruptcy moves AAV capsid IP into Lilly's lap, AbbVie pays $10.9B for Apogee, Remix takes RNA-modulator platform public

Sun, 28 Jun 2026 13:00:00 +0000

Portfolio Watch for the week ending Sunday, June 28, 2026. Three items move the thesis plus two macro framings. Item 1 — Sangamo Therapeutics Chapter 11 + asset sales (announced 2026-06-23). Chapter 11 filing in Delaware, DIP financing up to $30M, ~40% US workforce cut (51 of ~128 roles). Lilly stalking-horse bid: $50M + assumed liabilities for the capsid delivery platform, the zinc-finger platform, the modular integrase (MINT) platform, and the prion disease program ST-506. Astellas stalking-horse bid: up to $50M for Fabry program isaralgagene civaparvovec (ST-920). Section 363 sale — open to higher bidders at auction. Combined IP ceiling for one of biotech's most-cited gene-editing estates: $100M. Aletira read — bidirectional. Validation side: Lilly already owns Akouos (hereditary hearing-loss AAV gene therapy, Aletira's most natural strategic acquirer for the otology lead indication) and Verve base-editing assets from the 2025 acquisition. Lilly is now adding the Sangamo capsid library, ZFN platform, and MINT integrase on top — the same buyer that holds Aletira's commercial-validation logic on hearing-loss otology gene therapy is paying real (if depressed) money to stack delivery and engineering capabilities. The SELEXON premise — that cell-type selectivity is the next AAV problem to solve — is being underwritten by the player that owns Aletira's most likely acquirer slot. Threat side: Lilly inherits Sangamo's capsid engineering and capsid-evolution methods, which are the *alternative* path to tissue/cell-type selectivity. After this deal closes, the largest most-likely acquirer for an otology gene-therapy company will have a capsid-engineering arsenal in-house, and SELEXON's differentiation against capsid-engineering — splicing-based transgene-level regulation vs. capsid evolution — needs sharpening in the next BD pitch. Action items for Geoff Lynn at Aletira: (a) chase what is actually in the Sangamo capsid library — particularly any in-ear or CNS-tropic capsids the Lilly Akouos team will be able to repurpose; (b) make a courtesy outreach to the Lilly Akouos team to map the post-deal portfolio and start the BD relationship before integration completes, not after. Second-order read for the firm: Sangamo IP at $100M total recalibrates valuations downward across the broader AAV-delivery-and-engineering category — academic licensing terms should be more favorable for the rest of 2026, useful for Hemaka in JHTV / UM Ventures gene-therapy term negotiations. Item 2 — AbbVie acquires Apogee Therapeutics for $10.9B cash (announced 2026-06-22). Apogee: Cambridge MA clinical-stage antibody biotech. Lead APG777 (zumilokibart) is an extended-half-life IL-13 monoclonal in Phase 3 atopic dermatitis; APG808 is an IL-4Rα antibody (Dupixent target); APG273 is the IL-13/TSLP bispecific positioned against Sanofi-Regeneron amlitelimab and the broader TSLP wave. Year's #2 biotech buyout after GSK/Nuvalent ($10.6B). Strategic frame: post-Humira AbbVie immunology stack (Rinvoq + Skyrizi + ABBV-712) counterweight to Sanofi-Regeneron Dupixent encroachment. Read for Blackbird — not a direct portfolio touch. Indirect signals: (a) immunology M&A remains hot at double-digit-billion exit values for clinical-stage atopic-derm biology, which is the comp set backing exit valuation on the Slusher gut-restricted GCPII IBD program in adjacent inflammation-fibrosis biology; (b) Apogee was a Fairmount Funds / Atlas Venture biotech that IPO'd in 2023 at ~$300M and exits at $11B in three years — a 35x in three years on a clinical-stage immunology biotech under good conditions, a useful internal comp datapoint. Item 3 — Remix Therapeutics × Passage Bio reverse merger + $100M concurrent PIPE (announced 2026-06-24). All-stock reverse merger; Remix gets Passage's Nasdaq listing. Pre-merger Remix holders end with ~93%, Passage holders with ~7%. Closing Q4 2026. Remix lead REM-422 is an oral small-molecule mRNA degrader targeting MYB (historically undruggable TF). Ph1/2 at ASCO 2026: 43% ORR at the RP2D in biomarker-positive adenoid cystic carcinoma, 100% DCR, durable responses. ACC lead indication, AML second. PIPE leadership: Decheng Capital, Lynx1, Forge Life Science Partners. Two reads for Blackbird. (1) Remix degrades mRNA; Blackbird's portfolio RNA-platform asset upregulates mRNA translation — opposite direction, same conceptual neighborhood. A $100M PIPE into a fresh Nasdaq listing for an RNA-modulating small-molecule platform is real market validation for the category — tailwind for Avi when the Blackbird translation-activator program is positioned to outside investors. (2) Passage Bio's reverse-merger-out (after FDA pushback on their lead AAV gene-therapy asset, GM1 gangliosidosis) is the second restructuring/exit event in AAV gene therapy in two weeks (after Sangamo). The market has decompressed around generic AAV gene therapy and recompressed around RNA-modulating small molecules. Aletira positioning lesson: lean into the "SELEXON makes AAV gene therapy *more selective and therefore commercially viable*" frame in next investor pitch, not the generic "gene therapy is the future" frame. Position SELEXON as the *fix* for what the market is repricing. Macro framing 1 — capital is back. Week of 2026-06-22 to 2026-06-25 fresh equity tally: Ollin Biosciences $330M Series B (Vabysmo VEGF-A/Ang-2 bispecific challenger, OLN-324 in-licensed from Innovent, 3x Phase 3 in H2 2026); RQ Bio $115M Series A (long-acting flu prevention antibody RQB01, Christian Schade as exec chair); Lycia Therapeutics $75M Series D (LYTAC extracellular protein degraders, food allergy LCA-0061 + Graves' LCA-0321, Lilly in syndicate); Nura Bio $73.8M Series B (SARM1 inhibitors in ALS, Column Group + Sanofi Ventures); Boundless Bio + Serapha Bio merger with $230M PIPE (in vivo base editor for AATD, licensed from China's YolTech). ~$870M across 5 fresh-money deals in one week + the AbbVie/Apogee $10.9B M&A + the Sangamo $100M asset sale. Q3 2026 looks open for both fundraising and M&A on credible mechanism-first programs. Aletira, the GPR52 NewCo, the Slusher GCPII program if approaching readiness, and any Blackbird companies in or near financing windows should treat this as a green-light period. Macro framing 2 — the Lilly pattern. Last 18 months Lilly has acquired Akouos (otology gene therapy), Verve (base editing), 4E Therapeutics (chronic pain peripheral kinase, two weeks ago), and now the Sangamo gene-therapy estate at fire-sale price. Systematic in-house gene-editing + gene-therapy buildout — delivery, editing, disease programs all covered. Single most active strategic in the early gene-therapy and delivery space. Blackbird BD relationships with Lilly business development (not just the Akouos team) are an actionable item for Eddie. Lilly is paying for mechanism-first programs that fit the in-house platform thesis; the next two years continue that pattern. Closing actions. (1) Geoff Lynn (Aletira) — courtesy outreach to Lilly Akouos team before Sangamo integration completes; chase Sangamo capsid library specifics. (2) Hemaka — flex Sangamo's $100M IP price into JHTV/UM Ventures licensing negotiations; raise the Lieber-Blackbird platform conversation with Tom Hyde given Friday's + Sunday's Sourcing Radar Lieber asset stack. (3) Eddie — systematic BD relationship with Lilly business development beyond the existing Akouos contact. Sources surveyed: Fierce Biotech, Fierce Pharma, Endpoints, STAT, BioPharma Dive, GEN, BioBuzz, BusinessWire / GlobeNewswire / PR Newswire, JHTV / Hopkins Hub, Lieber Institute press, UM Ventures press. No source failures in window. No new direct portfolio-company press in window (Aletira / aSKY / Adventris / schizophrenia NewCo all quiet — typical for resident incubated companies pre-clinical readout). Already covered in prior weekly (2026-06-21 Portfolio Watch, excluded from this week): JHU Life Sciences Research Initiative ($80M/yr, 06-10/06-15); GSK/Nuvalent $10.6B (06-09); Neumora navacaprant Ph3 failure + layoffs (06-15); Lilly/4E Therapeutics chronic pain (06-16); Incyte/Vega VWD (06-08). Out-of-window mention from agent shortlist: JHTV 2026 Celebration of Innovation in Medicine + Slusher Dean's Distinguished Faculty Innovator Award — recap dated 2026-05-05, dropped as out of week-7-day window. Editorial call: Sangamo is the single most consequential item this week for the firm — direct bidirectional read on Aletira's strategic acquirer profile and on platform-IP valuations broadly. Apogee deal anchors immunology exit comps; Remix listing anchors RNA-modulator category validation and Aletira positioning narrative. Macro environment + Lilly pattern give the action frame.

Sourcing Radar — Hopkins mKRAS-VAX with ipi/nivo hits primary endpoints in metastatic MMR-proficient colorectal cancer

Sat, 27 Jun 2026 12:00:00 +0000

Sourcing Radar for Saturday, June 27, 2026. Lead. Wang HH, Huff AL, Haldar SD, Berg M, Thoburn C, Heumann TR, Anders RA, Barrett B, Bever KM, Pishvaian MJ, Lee V, Le DT, Azad NS, Yarchoan M, Zaidi N, Jaffee EM (senior, corresponding, Sidney Kimmel Comprehensive Cancer Center + Johns Hopkins Convergence Institute + Bloomberg-Kimmel Institute for Cancer Immunotherapy, Johns Hopkins). "Mutant KRAS peptide vaccine with dual checkpoint blockade in metastatic colorectal cancer: a phase I trial." Nature Communications, posted 2026-06-23, DOI 10.1038/s41467-026-74711-8. Second readout out of NCT04117087 — the metastatic MMR-proficient / MSS colorectal-cancer arm of the same pooled mutant-KRAS synthetic-long-peptide-vaccine program whose resected-pancreatic arm published in February (DOI 10.1038/s41467-026-68324-4). Design. mKRAS-VAX — one off-the-shelf product pooling six 21-mer SLPs covering the common KRAS hotspot mutations G12D, G12V, G12C, G12R, G12A, G13D — adjuvanted with poly-ICLC (Hiltonol, Oncovir), combined with ipilimumab + nivolumab during weekly-times-four priming, followed by nivolumab maintenance plus vaccine boosters every 8 weeks. Same vaccine product regardless of which KRAS allele a patient carries. Results — metastatic CRC arm (this paper, N=13 pretreated metastatic MMR-p / MSS CRC). Both primary endpoints (safety, immunogenicity within 17 weeks of vaccination) met. All vaccine-attributable AEs grade 1-2; no incremental severe irAEs beyond the dual-CPI baseline profile. Immunogenicity: mKRAS-reactive T cells expanded in 8/12 (75%) biomarker-evaluable patients by direct ex vivo IFN-γ ELISpot, and in 12/12 (100%) following in vitro expansion. Secondary RECIST 1.1 efficacy endpoints reported as supportive; full ORR/DCR/PFS/OS in the supplement (not in the abstract). Companion data — resected PDAC arm (Feb 2026 Nat Comm, same trial): N=12, 11/12 immune responders, median DFS-from-vax 6.35 months, median OS 29.59 months, 4/12 disease-free at last follow-up (all G12V or G12R); top-quartile immune responders DFS 18.8 months vs. 2.76 months bottom quartile. Strongest correlate of DFS across both arms is the patient's own mKRAS-reactive T-cell response, not baseline tumor or patient features. Why it matters. MSS / MMR-proficient KRAS-mutant metastatic CRC is roughly the worst-prognosis solid-tumor subgroup and the canonical immune-checkpoint-cold population; ~40% of metastatic CRCs carry a KRAS hotspot mutation. Off-the-shelf single-product six-mutation coverage is the technical bet. Causal logic on a single-arm Phase I is provisional, but immunogenicity-to-DFS correlation replicating across two tumor types in the same trial is real signal. Blackbird angle — three reads. (1) Competitive landscape, with direct aSKY read-through. The vaccine is licensed from Johns Hopkins to Adventris Pharmaceuticals (Y Combinator-backed; Liz Jaffee co-founder + CSO; Mark Yarchoan and Neeha Zaidi hold equity). Adventris is the Hopkins commercialization vehicle for the mKRAS-VAX program and the broader Jaffee-lab neoantigen-vaccine IP. Field leader on readouts is Elicio Therapeutics (Cambridge, MA) — ELI-002, an amphiphile-conjugate KRAS vaccine that traffics to lymph nodes via endogenous albumin binding. ELI-002 2P: 21/25 (84%) immune response in AMPLIFY-201 PDAC + CRC MRD. ELI-002 7P: 99% in AMPLIFY-7P PDAC MRD; Phase 2 fully enrolled (135 patients) and is the next industry-defining readout in this space. Different mechanistic bets — Hopkins/Adventris is adjuvant + dual CPI; Elicio is lymph-node delivery without CPI — and both can win different sub-indications. (2) aSKY portfolio read-through. aSKY's disclosed AVS Bio work is a KRAS-frame antibody-engineering effort, not a peptide vaccine; specific MOA still nondisclosed. The strategic differentiation question for Krupnick's team: with Elicio and Adventris about to define peptide-vaccine standards across MRD and metastatic CRC/PDAC, what sub-segment best fits an antibody-engineered KRAS-recognition approach (and how does it position vs. precedent including the Xyphos/Convergence body of work and the Hopkins-Berkeley pMHC mutant-KRAS antibody)? Hemaka should surface this question to the aSKY team. (3) Adventris itself as a sourcing question. Y-C-backed Hopkins-licensed Jaffee-founded with two Phase 1 readouts now public — this is squarely a company Blackbird BioVentures should have a position on (invested, passed with thesis, or actively monitoring). Eddie should confirm the JHTV deal structure on Adventris and whether the Convergence Institute relationship affords any preferred-access angle on follow-on Jaffee-lab neoantigen-vaccine programs (mutant TP53, mutant EGFR, broader pan-cancer driver-mutation neoantigen lane). Tail item. Danilova L, Favorov A, Smith KN, Cope L (corresponding: Ludmila Danilova, JHU School of Medicine). bioRxiv DOI 10.64898/2026.06.18.733036, posted 2026-06-23. replicateFest — R package + Shiny app for analysis of T-cell-receptor-repertoire data from FEST (Functional Expansion of Specific T cell) assays, the workhorse readout for identifying neoantigen-specific T-cell clonotypes that underpins every immunogenicity readout in the neoantigen-vaccine space — including mKRAS-VAX. Methods/software, not a sourcing lead. The reason it's on the radar: Kellie Smith at Bloomberg-Kimmel runs one of the deepest neoantigen-TCR-discovery groups in academia, and her tooling generally precedes her therapeutic concepts by 12-24 months — a useful future-watch name for the next personalized-neoantigen-TCR concept out of Hopkins. Recommended actions. (1) Eddie to confirm JHTV deal structure on Adventris and any preferred-access angle on follow-on Jaffee-lab neoantigen-vaccine programs via the Convergence Institute relationship. (2) Hemaka to surface to aSKY the modality-differentiation question against the now-credible Elicio + Adventris peptide-vaccine standards in MRD and metastatic CRC/PDAC. (3) Internal note: track Adventris's next readout cadence; add Kellie Smith to the JHU future-watch list. Candidate funnel — 2026-06-24 to 2026-06-27. PubMed (Johns Hopkins/UMB/Lieber affiliation × drug discovery / therapeutic target / mechanism, 30 returned, 23 metadata-fetched): Hopkins Wang/Jaffee Nat Comm mKRAS-VAX metastatic CRC Phase 1 (CHOSEN as lead); Hopkins Giraldo/McMeniman PNAS insect-fungal scent/adhesion (basic biology, not translational); Hopkins Hauk/Berger PNAS DCTPP1 antagonists + decitabine synergy (shipped 2026-06-21); Hopkins Morakis/Durr Blood Adv sickle cell OBM (shipped 2026-06-23); Hopkins/NIDA-IRP Jahan/Newman Eur J Med Chem etonitazene dual MOR-D3 ligand (NIDA-IRP lead institution, Rana Rais of JHU Drug Discovery a contributor only, not a sourcing lead); Hopkins Cosgrove ABSSSI DOOR endpoint (clinical-trial methods, not therapeutic); JHU/SKCC Anders et al. KRAS vaccine CRC (= same paper, primary); JHU/SKCC DeZern in MDS taxonomy validation (multi-institution consortium, not lead). Other PubMed: most non-JHU/UMB/Lieber. bioRxiv (search_preprints 2026-06-23 to 2026-06-27, categories pharmacology + cancer biology + neuroscience + bioengineering + immunology + genetics + cell biology; ~150+ items scanned, JHU/UMB/Lieber-affiliated items checked individually): Camilo-Contreras/Simonetti HIV ag qVOA v2 (shipped 2026-06-24); Danilova/Smith/Cope replicateFest TCR (CHOSEN as tail). No fresh UMB/Lieber-anchored translational item in window. Institutional press (hub.jhu.edu, hopkinsmedicine.org, libd.org, ventures.jhu.edu, umventures.org, medschool.umaryland.edu via WebSearch + r.jina.ai mirror; hub.jhu.edu and hopkinsmedicine.org 403 direct from this IP): nothing fresh portfolio-touching in window — JHU 2026 announcements that surfaced are out of window (June 10 Life Sciences Research Initiative launch, Syngene/JHU deal Feb 2026). Industry press (BioBuzz, Fierce, Endpoints, BioPharma Dive, STAT): nothing portfolio-touching that shifts today's call. Sources that failed or were proxied: hub.jhu.edu and hopkinsmedicine.org (Cloudflare/403, used r.jina.ai mirror + WebSearch snippets); www.nature.com/articles/s41467-026-74711-8 direct fetch (302 to idp.nature.com auth wall, full text via r.jina.ai mirror partial — abstract + acknowledgements only; supplement not extractable). Already shipped this week (excluded): DCTPP1/decitabine (06-21), Wilmer crystal sorafenib AMD (06-22), epigenetic vaccine sarcoma (06-23), OBM sickle-cell vaso-occlusion (06-23), agQVOA HIV deep latency (06-24), Lieber trans-eQTL INGENE/MODULE schizophrenia targets (06-25), Lieber sign-tracker/goal-tracker fMRI D2 biomarker + JHU LipoGlo (06-26). Editorial call: paper is in-window (NCT04117087 metastatic CRC arm public Nat Comm 2026-06-23) and is the cleanest commercial-implication item for Blackbird in the window — Hopkins-licensed program, Hopkins-spinout commercialization vehicle (Adventris), direct read-through to aSKY portfolio modality choice. Paper link: https://www.nature.com/articles/s41467-026-74711-8

Sourcing Radar — Lieber's sign-tracker/goal-tracker fMRI biomarker stratifies D2 antipsychotic response

Fri, 26 Jun 2026 12:00:00 +0000

Sourcing Radar for Friday, June 26, 2026. Lead. Sambuco, Lupo, Hawkins, Selvaggi, Antonucci, Bertolino, Blasi, ... Howes O.D. (KCL), Mehta M.A. (KCL), Pergola G. (corresponding, Lieber Institute for Brain Development) — bioRxiv v3 posted 2026-06-24 (10.64898/2026.05.14.724267), "Reproducible Human Reward Imaging Phenotypes Exhibit Differential Sensitivity to Dopamine D2 Receptor Antagonism." Translates the rodent sign-tracker (ST) / goal-tracker (GT) framework to humans using fMRI BOLD on the Monetary Incentive Delay (MID) task, clustering on the anticipation-vs-outcome activation profile. Two-cluster solution recovers cleanly across 5 cohorts (N=1,251 total: Discovery IMAGEN N=890; Replication Bari N=245; single-dose pharma N=34; 7-day repeated-dose pharma N=48; clinical FEP/SCZ N=34) with replication accuracies 77-94%. ST: heightened ventral striatal BOLD during reward anticipation; GT: inverse pattern, heightened BOLD during outcome. Pharmacology — the operative finding. Single-dose risperidone (0.5/2mg, dose-linear) + haloperidol (3mg), full D2/D3 antagonists, selectively suppress ST anticipatory ventral striatum BOLD (Drug x Cluster F(1,30)=11.58, p=.002, eta²p=.279), with parallel decrease in self-reported Energy on VAS only in ST (F(1,30)=20.47, p<.001, eta²p=.41). 7-day amisulpride (full D2/D3 antagonist) replicates in the repeated-dose cohort (Treatment x Cluster F(1,37)=5.69, p=.022) — critically, recovers a real pharmacological effect that was reported null in the group-level analysis of the same dataset (Osugo et al.), a textbook biomarker-enabled drug-effect rescue. 7-day aripiprazole (D2/D3 partial agonist) produces opposite-sign effects: increased anticipatory BOLD in GT (β=0.86, p=.023), decreased outcome BOLD in GT (β=-1.07, p=.021), no change in ST. Cross-compound divergence rules out regression-to-the-mean. Clinical cohort (UNIBA, N=34 FEP+SCZ on standing antipsychotics): ST patients show blunted anticipatory ventral striatum vs. healthy ST (d=2.06); patient-specific D2 dopaminergic affinity index correlates negatively with anticipatory BOLD (r=-0.462, p=.023) and positively with PANSS Negative symptoms (r=+0.471, p=.020) — specifically D2 (D1, D3, D4, D5 all null). GT patients have significantly higher negative symptom burden than ST patients on the same medications. Blackbird angle — three reads. (1) Direct pull-through to the GPR52 NewCo with Lieber and Third Rock. The published rationale for the "go around D2" thesis is now the cleanest single paper in the field, and it comes out of Lieber. Forward-looking utility is trial-design tool: enrich the GPR52 POC on GT-phenotype patients identified by this imaging biomarker — patients with high baseline negative-symptom burden whose anticipatory reward circuit is not D2-driven — to produce a clean signal for a non-dopaminergic mechanism against a muted comparator-class signal. That is a study-power and trial-cost frame Esther should run past Third Rock. (2) Platform read. Pergola has published two adjacent Lieber infrastructure assets inside a month — the trans-TWAS (INGENE/MODULE) target-discovery engine covered Thursday, and now this fMRI patient-stratification tool on the development side. The question for Hemaka in the next Tom Hyde conversation is whether the broader Lieber neuropsychiatry platform stack (postmortem transcriptomics + MRI phenotyping + iPSC/astrocyte work) adds up to a licensable platform-layer asset underwriting multiple single-asset spinouts — the platform-layer next-order question following the GPR52 single-asset proof. (3) Candor flag — biomarker correlation paper, not prospective trial; clinical sample small (26 concordant after exclusions) and Bari-only. Pull-through question for Pergola directly: open to a prospective biomarker-validation arm alongside a Lieber-Blackbird industry partner's clinical program? Second item. Kelpsch, Zhang, Thierer J.H., Koren, Kumar, Lin, Hensley, Sohn, Liu J.O., Lectka, Mumm J.S., Farber S.A. (corresponding, Johns Hopkins University) — bioRxiv v4 posted 2026-06-24 (10.1101/2024.11.14.623618), "A whole-animal phenotypic drug screen identifies suppressors of atherogenic lipoproteins." LipoGlo zebrafish reporter (NanoLuciferase on ApoB) used to screen ~3000 compounds of the Johns Hopkins Drug Library (JHDL) on the ARQiv-HTS platform — 4 doses x 8 replicates each, ~55,000 individual measurements, MTP inhibitor lomitapide as positive control on every plate. Stringent SSMD + fold-change cutoffs yield 49 unique B-lipoprotein-lowering hits; 19 validate orthogonally; 7 extensively phenotyped. Enoxolone (licorice root) selected and demonstrated via clean genetic-rescue logic to lower B-lps through HNF4α — effect lost in HNF4α mutants, reproduced by independent HNF4α inhibitors. Commercial angle. ApoB-lowering market is the largest therapeutic market in the world, still dominated by ApoB-synthesis (lomitapide, mipomersen) or clearance (statins, PCSK9, bempedoic acid) mechanisms; statins leave ~30% residual CV mortality; the newer mechanisms are priced for severe dyslipidemia only. Vertebrate phenotypic screens at scale are rare — the discovery infrastructure is what is missing. Farber lab demonstrates with this paper: (i) the screening platform works at scale and recovers the positive control with quantified hit-rate; (ii) mechanism-of-action triage is possible directly in the vertebrate system using genetic-rescue logic — compresses the MoA-to-target loop; (iii) the hits include a previously unannotated nuclear-receptor mechanism (HNF4α) with a small-molecule starting point. Team is the JHU translational stack: Farber (senior), Joseph Thierer (recently spun out into independent leadership on the ApoB program), Jeff Mumm (built ARQiv-HTS), Jun O. Liu (curates JHDL, deep track record of high-throughput drug-repositioning to companies). Blackbird angle for Avi and Hemaka: lead is not the ApoB program itself (3-4 credible companies already in cardiometabolic ApoB space, Hopkins will be testing strategic-partner waters) — lead is the LipoGlo platform as a discovery service or platform spinout. Vertebrate phenotypic screening for lipid biology generalizes to NASH, lipodystrophy, familial chylomicronemia, fatty liver, and the cardiometabolic-renal axis where SGLT2/GLP-1 franchises are sitting on the largest budgets in industry. Differentiation: whole-animal vertebrate readout for a class of biology where cell-based screens systematically miss tissue-level lipid trafficking. Candor caveat: platform spinouts in screening are notoriously hard to value. Recommended action: JHTV-routed conversation with Farber group probing whether Joseph Thierer is interested in operator leadership and whether the HNF4α small molecule has composition-of-matter coverage independent of platform claims. Recommended actions (consolidated): (1) Esther to Third Rock on biomarker-enrichment frame for the GPR52 NewCo clinical plan. (2) Hemaka to Tom Hyde at Lieber on the broader Lieber neuropsychiatry platform stack as licensable infrastructure beyond the single-asset GPR52 deal. (3) Avi/Hemaka JHTV-routed read on the LipoGlo platform (not the ApoB program) — Thierer leadership availability, HNF4α small-molecule composition-of-matter status. Candidate funnel — 2026-06-22 to 2026-06-26. bioRxiv (api.biorxiv.org details API, paginated 0-3000 across 2026-06-19 to 2026-06-26, full sweep filtered on Hopkins / Lieber / UMB / Wilmer / Kennedy Krieger / Bloomberg School / Kimmel corresponding institutions — 4 in-window hits): Pergola Lieber reward imaging v3 (CHOSEN as lead); Farber JHU LipoGlo v4 (CHOSEN as second); Salzberg JHU EviAnn genome annotation v3 (10.1101/2025.05.07.652745, bioinformatics tool, no IP, dropped); Danilova/Smith/Cope JHU SoM replicateFest TCR R package (10.64898/2026.06.18.733036, methods/software, dropped). medRxiv (paginated 0-2000): Baughman JHU EHR-integrated GenAI VTE risk stratification (10.64898/2026.06.17.26355819, clinical AI, low translational-IP fit, dropped); Goyal/Stevens JHU deep-learning AIS detection on NCCT (10.64898/2026.06.20.26356152, clinical AI, dropped); JHU Neurology subtle aphasia post-stroke (10.64898/2026.06.19.26356022, observational, dropped). Institutional press (hub.jhu.edu, hopkinsmedicine.org, libd.org, ventures.jhu.edu, umventures.org, medschool.umaryland.edu): hub.jhu.edu and hopkinsmedicine.org return 403 from this IP (Cloudflare), proxied via r.jina.ai — no fresh translational spinout / licensing / preclinical-result items in window. UM Ventures most recent translational press 2026-06-09 (out of window). Industry press (BioBuzz, Fierce Biotech, Endpoints, BioPharma Dive, STAT): nothing portfolio-touching that shifts today's call. Sources that failed or were proxied: hub.jhu.edu (403, Jina proxy); hopkinsmedicine.org (403, Jina proxy). Already shipped this week (excluded): DCTPP1/decitabine (06-21), Wilmer crystal sorafenib AMD (06-22), epigenetic vaccine sarcoma (06-23), OBM sickle-cell vaso-occlusion (06-23), agQVOA HIV deep latency (06-24), Lieber trans-eQTL INGENE/MODULE schizophrenia targets (06-25). Editorial call: Pergola paper is a v3 with substantial new pharmacological + clinical cohort additions vs. v1 abstract — treated as in-window publication. Farber LipoGlo is v4 (original v1 Nov 2024) but v4 lockdown this week is the canonical form; included as second item under the same in-window-revision logic that surfaced the agQVOA v2 on Wednesday. Paper link: https://www.biorxiv.org/content/10.64898/2026.05.14.724267v3

Sourcing Radar — Lieber's trans-eQTL engine surfaces 641 new schizophrenia genes

Thu, 25 Jun 2026 12:00:00 +0000

Sourcing Radar for Thursday, June 25, 2026. Lead. Giulio Pergola (corresponding) and Daniel Weinberger (senior, Lieber Institute CEO/director), in Nature Genetics, June 22, 2026 — "Co-expression-based models improve eQTL predictions for transcriptome-wide association studies and highlight new schizophrenia-associated genes." They build two trans-aware predictive models on top of Lieber Institute (LIBD) postmortem brain transcriptomics across six regions (DLPFC, hippocampus, caudate, amygdala, dACC, sACC). INGENE models how cis-regulated coexpression partners of a target gene drive that gene through trans effects. MODULE associates SNPs with the eigengene (first principal component) of a target gene's coexpression module. Both use elastic-net regression with cross-validation; both require replication across LIBD, GTEx, and CommonMind to filter out spurious trans signal — the filter that distinguishes this from prior trans-TWAS attempts that did not hold up. Application. Trained models applied to PGC3 SCZ GWAS genotypes from 102,613 individuals (62 cohorts) identify 766 FDR-significant SCZ-associated genes — 641 of them novel transcriptome-wide. Cross-cohort concordance ~69%. INGENE yields ~1.8x more predicted genes than CIS-only models and ~1.7x more than EpiXcan; GWAS effect-size correlation roughly doubles vs. cis (r=0.28–0.42 trans vs 0.14–0.18 cis). Biology converges on AMPA receptor trafficking and synapse organization in DLPFC; antigen-processing/MHC pathways in prefrontal and cingulate cortex; cell-type-resolved signal shows pronounced excitatory-neuron disease-associated expression with GABAergic-interneuron downregulation in DLPFC (the excitation/inhibition imbalance model, with a longer pursuable target list). Top replicable hits: MAPK3 (glutamatergic-synaptic plasticity), GATAD2A (NuRD chromatin-remodeling), INO80E, GNL3, SNX19, CYP21A1P, C4A. Limitations the authors acknowledge: bulk postmortem tissue (cell-type specificity blurred), European-ancestry restricted (PGC3), no sex-effect modeling, statistical predictions not causal claims. Blackbird angle — two reads. (1) Platform/infrastructure read. Pergola's group has built and demonstrated in print a target-discovery engine that converts LIBD's unique postmortem cohort into a high-quality novel-target shortlist for schizophrenia — i.e., infrastructure Lieber owns. Lieber's existing Boehringer Ingelheim relationship is precedent that pharma counterparties pay for this kind of asset. Pull-through: conversation with Tom Hyde and Dan Weinberger on whether INGENE/MODULE is operationalizable as licensable infrastructure beneath a series of LIBD single-asset spinouts (the existing GPR52 program with Third Rock is the single-asset proof; this is the platform-layer next question). (2) Gene-list read. The 641 novel SCZ-associated genes intersect modalities Blackbird's molecular discovery team can pursue — AMPA-trafficking machinery, cortical antigen-processing nodes, MAPK3, chromatin-remodeling factors. Worth Avi running the list through internal druggability filters (small-molecule, antibody, splice-modulation tractability) and surfacing the intersection with Hemaka's LIBD relationships. Caveat: trans-TWAS is statistical, not causal — triage matters before this becomes a pipeline. Supporting item — freshness disclosed. Shreesh Mysore (corresponding, JHU Psychological & Brain Sciences), Nature Communications, April 28, 2026 — "Evolutionarily old brainstem neurons are required for the control of selective spatial attention." (JHU press push 2026-06-22; the paper is older, the press is new — flagged candidly in the script.) Identifies PLTi as a conserved brainstem inhibitory region that gates selective spatial attention by controlling competing-stimulus representations in the superior colliculus. Bilateral silencing in freely behaving mice severely disrupts target selection without breaking perception or motor execution — a clean attention-specific phenotype consistent with ADHD-like distractibility. First-in-class circuit-level attention target; mechanistically distinct from monoamine-targeted ADHD therapy. Modality-open IP runway: no annotated receptor or ligand here yet, but the anatomical target is novel. Action: future-watch on the Mysore lab; not a meeting this week, but a name to know. Recommended actions. (1) Hemaka/Esther conversation with Tom Hyde and Dan Weinberger at Lieber on INGENE/MODULE as licensable platform infrastructure beyond single-asset deals. (2) Avi to run the 641-gene list through internal druggability filters and intersect with Blackbird-tractable modalities. (3) Add Shreesh Mysore (JHU Psych & Brain Sciences) to the future-watch sourcing list. Candidate funnel — 2026-06-22 to 2026-06-25. bioRxiv (api.biorxiv.org, paginated 0-500, 162 total in-window; transient 503s resolved on retry): only JHU/UMB/Lieber hit was the Camilo-Contreras ag qVOA preprint v2 (10.64898/2026.06.11.731680) already shipped 2026-06-24. medRxiv (paginated 0-300, 312 total): Baughman et al. multisite EHR-integrated GenAI VTE risk stratification (JHU contributing, 10.64898/2026.06.17.26355819) — clinical AI deployment, low translational fit, dropped. Journals: Pergola/Weinberger Nat Genet June 22 — CHOSEN as lead; Mysore Nat Commun Apr 28 — CHOSEN as supporting (paper out of window, JHU press in window, disclosed); Feinberg/Hansen/Threadgill Nat Genet May 20 (Hub press June 23) — paper substantially out of window, passed. Institutional press: hub.jhu.edu and hopkinsmedicine.org 403 direct from this IP (routed via r.jina.ai mirror and WebSearch snippets); libd.org no June press item; UM Ventures and JHTV no fresh spinout news in window. Ecosystem (technical.ly, business.maryland.gov): Maryland Future 20 named Vita Therapeutics and Microbiome Foundries (JHU spinouts) — context, not Radar picks. Industry press (BioBuzz, Fierce Biotech, Endpoints, BioPharma Dive): JHU + Revolution Medicines RAS(ON) combo data trail at ESMO GI (JHU site role only), no portfolio-relevant signal. Sources that failed (or were proxied): hub.jhu.edu and hopkinsmedicine.org (Cloudflare/403, used r.jina.ai), GEN article URL (404, used EurekAlert/Medical Xpress/Nature.com instead). Already shipped this week (excluded): DCTPP1/decitabine (06-21), Wilmer crystal sorafenib AMD (06-22), epigenetic vaccine sarcoma (06-23), OBM sickle-cell vaso-occlusion (06-23), ag-qVOA HIV deep latency (06-24). Paper link: https://www.nature.com/articles/s41588-026-02646-3

Sourcing Radar — Hopkins ag-qVOA reactivates HIV reservoirs from deep latency

Wed, 24 Jun 2026 12:00:00 +0000

Sourcing Radar for Wednesday, June 24, 2026 — a quiet bioRxiv window for clean Hopkins-led sourcing, with one strong lead and one tool note. Lead. Camilo-Contreras, Simonetti (corresponding, Johns Hopkins School of Medicine, Department of Medicine — Division of Infectious Diseases), and the Siliciano group, with Wistar/Penn collaborators (bioRxiv DOI 10.64898/2026.06.11.731680, v2 posted to JHU bioRxiv channel 2026-06-23; v1 2026-06-12). They build an antigen-restricted quantitative viral outgrowth assay (ag qVOA) that swaps the standard mitogen (PHA) for cognate antigen-driven TCR stimulation, then test it on two extensively-characterized PWH whose dominant intact proviruses sit in "deeper latency" chromatin: P012 carries a pericentromeric (PeriC) provirus on Chr14 in a Candida albicans-responsive expanded clone, and elite controller ES24 carries a ZNF721i provirus in an HIV-1 Gag(P61)-reactive clone. CA lysate stimulation drove exponential outgrowth from P012's PeriC provirus (100% of recovered env sequences); P61 peptide drove outgrowth from ES24's ZNF721i provirus that PHA could not induce at all (~10x higher inducibility than the std. qVOA, all 22 U5-gag sequences matched ZNF721i). PeriC was independently detectable in P012's residual plasma viremia under suppressive ART (2/17 plasma env sequences matched), confirming the deep-latency reservoir is biologically active in vivo. So-what. Recasts the HIV-1 cure roadmap. The "deeper latency" pool the field had largely written off as inert is in fact reactivatable through the same physiological signal (cognate antigen on TCR) that probably drives in vivo reactivation in patients — and that signal can hit chromatin contexts the standard LRA toolkit (HDAC inhibitors, PKC agonists) can't reach. Validates therapeutic vaccination and TCR-engager strategies (both pulling reservoir into expression for clearance) over the pharmacologic shock-and-kill paradigm that has stalled clinically across the field for fifteen years. The ag qVOA itself is a defensible platform asset — better than std. qVOA as a diligence assay for cure trials because it measures the reservoir population that actually behaves like the in-vivo reservoir. Blackbird angle. Hopkins owns the HIV latency franchise — Bob Siliciano, Janet Siliciano, Joel Blankson, Stuart Ray, and rising-figure Francesco Simonetti (corresponding) — in a way no other academic center does. HIV cure is a high-stakes, well-funded space (Beat-HIV and I4C Martin Delaney collaboratories alone are several hundred million in NIH cycles, plus strategic interest from Gilead/ViiV/Merck on the post-lenacapavir question). Pull-through: JHTV-routed meeting with Simonetti on (a) IP filing status around the ag qVOA platform, (b) whether the assay is spinnable independently from a therapeutic cure vehicle, (c) whether the Siliciano lab has companion therapeutic IP (e.g., antigen-driven killing or TCR-engager constructs) that could co-launch. Bob Siliciano as natural advisor seat. Tool note. Yovanno & Lau (Johns Hopkins School of Medicine — Department of Biophysics & Biophysical Chemistry), bioRxiv DOI 10.64898/2026.06.22.733418, posted 2026-06-23. TEntroPy — an open-source Python library that builds directional protein allostery networks from molecular dynamics trajectories using transfer entropy, identifying broadcaster and receiver residues and optimal information-flow paths between orthosteric and allosteric binding sites; perturbation of key binding-site residues changes the TE-weighted network in ways that trace communication routes between sites. No IP play, so not a sourcing lead — but the use case for Blackbird's molecular discovery team is direct. The GPR52 agonist program with Lieber Institute is exactly the kind of allosteric GPCR where understanding directional information flow between the orthosteric and allosteric pockets is the thing that distinguishes a developable molecule from a non-developable one. Worth flagging to Avi if the GPR52 SAR campaign is at a point where mapping the allosteric network of the lead chemotype would change the design loop. Recommended actions. (1) JHTV-routed meeting with Francesco Simonetti (and Bob Siliciano as advisor seat) on ag qVOA platform IP, next-cohort plans, and whether the assay spins independently from a therapeutic cure vehicle. (2) Internal note to Avi Khanna: TEntroPy as a diligence tool for the GPR52 allosteric SAR work with Lieber. Candidate funnel — last 3-4 days. bioRxiv JHU channel (2026-06-22 to 2026-06-24, 11 most-recent items): Siliciano ag qVOA (CHOSEN as lead); Yovanno/Lau TEntroPy allostery (CHOSEN as tool note); Carranza/Krakauer/Wittenberg cervical SCS for stroke arm rehab — Reach Neuro (Pitt) IP per the competing-interests statement (founders Weber, Capogrosso, Gerszten), so not a JHU/UMB sourcing lead; framed and dropped as competitor watch; Goyal/Stevens deep-learning ischemic-stroke detection on non-contrast CT (medRxiv, abstract-only render via Jina, dropped); Baughman et al. multisite EHR-integrated GenAI VTE risk stratification (clinical AI deployment, JHU contributing authors only, low Blackbird translational fit); Danilova/Smith/Cope replicateFest TCR repertoire R package (computational, not commercial); PVDOMICS pulmonary-hypertension lipidomics (cohort study, not platformable); plus several non-translational items. Out-of-window but worth scoping: Markham/Sears (JHU SoM/Bloomberg, second-to-last)/Coffey (Vanderbilt, senior) early-stage CRC spatial atlas (bioRxiv DOI 10.64898/2026.06.18.733268, 6 days back) — Hopkins contribution worth a separate Sears-pipeline conversation but pulled from today's episode because the bioRxiv details API hasn't yet indexed the DOI (page is reachable via Jina mirror, sync lag); pediatric sarcoma epigenetic GVAX (Recho/Ladle) shipped 2026-06-23. UMB/Lieber bioRxiv channels: no fresh items in window matching translational criteria. Institutional news (Hopkins Medicine newsroom, JHTV news, Lieber news, UM Ventures): no June 22-24 items beyond what's already preprinted. Press feeds: nothing new on portfolio competitors. Paper link: https://www.biorxiv.org/content/10.64898/2026.06.11.731680v2

Sourcing Radar — Hopkins ag-qVOA reactivates HIV reservoirs from deep latency

Wed, 24 Jun 2026 12:00:00 +0000

Sourcing Radar — Hopkins ag-qVOA reactivates HIV reservoirs from deep latency

Wed, 24 Jun 2026 12:00:00 +0000

Sourcing Radar — Hopkins wakes immune-cold pediatric sarcomas with an epigenetic GVAX

Tue, 23 Jun 2026 22:00:00 +0000

Bonus Sourcing Radar for the back half of June 23, 2026. Single-item episode on a new bioRxiv preprint from Brian H. Ladle (corresponding) and Michael A. Koldobskiy at the Johns Hopkins Sidney Kimmel Comprehensive Cancer Center, with Nicolas J. Llosa and the JHU pediatric oncology group (Recho, Gatla, Resch, Phillips, Glavaris, Doucet, Looi, Barbato, Llosa, Koldobskiy, Ladle; DOI 10.64898/2026.06.18.733244; posted June 18, 2026, v1). The team uses sequential decitabine (hypomethylating agent) followed by entinostat (HDAC1/3 inhibitor) to drive broad transcriptional reactivation of epigenetically silenced genes in a Kras-p53 murine sarcoma model (KP Sarc), including cancer testis antigens and a ~16x increase in surface MHC class I. Critically, irradiated epigenetically-treated KP Sarc cells mixed with a GM-CSF-secreting bystander line (GVAX) vaccinate mice against subsequent challenge with epigenetically-treated KP Sarc: 85% complete rejection when combined with anti-PD-1 + anti-CTLA-4 ICI, T cell-dependent (CD4+/CD8+ depletion ablates), long-term memory at 60-day rechallenge, and — the key cross-tumor result — protection against an unrelated syngeneic sarcoma (M3-9-M) treated with the same epigenetic regimen, demonstrating shared epigenetically-induced antigens. Validated transcriptomically in three human pediatric sarcoma lines (RH30, RH41 PAX3::FOXO1 RMS; CHLA-9 EWSR1::FLI1 Ewing) with 240 shared upregulated genes including MAGEA4 (Adaptimmune Tecelra target), PRAME (Immatics IMA203), and LRRC15. Blackbird angle: (a) acute unmet need in pediatric sarcoma — survival hasn't moved in 30 years, ~1,500 US patients/year, strong FDA pediatric/orphan/RPDD pull, Adaptimmune Tecelra accelerated approval (Dec 2024) is the live precedent; (b) differentiated mechanism — polyclonal vaccine response across a CTA cocktail, rather than single-antigen T-cell therapy or ADC, harder to escape; (c) IP layered cleanly — molecules are off-patent (de-risked) but the sequential-priming-plus-vaccine method is process-patentable and the broader platform claim (epigenetic priming for low-mutation tumors) supports a biotech-scale check; (d) heritage — GVAX was invented at Hopkins by Pardoll and Jaffee in the 1990s, became Aduro Biotech (then Chinook → Novartis); original GVAX programs failed Ph3 in pancreas/prostate using GM-CSF adjuvant alone, but this work re-engineers the bet by enriching the antigen substrate via epigenetic priming; (e) team — Ladle (peds onc immunotherapy), Koldobskiy (epigenetics), Llosa (sarcoma clinical) at Sidney Kimmel, with Pardoll as natural advisor and JHU COG network for pediatric Ph1. Honest read: mouse-model data on inbred BL/6 background; human cell-line data encouraging but cell lines aren't patients; centralized allogeneic manufacturing required (the Aduro path) versus unscalable autologous; entinostat carries E2112 Ph3 baggage from Syndax (though mechanism here is independent); Ladle has not been founding CEO of a venture-backed company. Recommended action: JHTV-routed meeting with Ladle, Koldobskiy, Llosa, and Pardoll on IP family breadth, the IND-enabling preclinical package (PDX + NSG models, allogeneic vaccine cell-line manufacturing), and whether a COG-anchored Phase 1 in relapsed pediatric sarcoma is the right first clinical readout — then a Blackbird translational grant into the IND package with BioHub as incubation home. Source: bioRxiv 10.64898/2026.06.18.733244, full text via r.jina.ai mirror.

Sourcing Radar — Hopkins BME images sickle-cell vaso-occlusion live in patients

Tue, 23 Jun 2026 12:00:00 +0000

Daily Sourcing Radar for June 23, 2026. Lead pick is a Blood Advances paper from Nick Durr's biomedical engineering lab at Johns Hopkins, with Lydia Pecker (Hopkins Hematology) on the clinical side (Morakis, Huang, McKay, Lanzkron, Pecker, Durr; DOI 10.1182/bloodadvances.2025018716; PMID 41779978; PMC13273109; vol 10 issue 12 pp 4215-4226; published June 23, 2026). The team used oblique back-illumination microscopy (OBM) — a label-free, noninvasive optical technique — to image ~91 sublingual capillaries in each of 10 sickle cell patients before and after RBC transfusion, against 10 unaffected controls. They directly visualize sickled RBCs adhering to the endothelium and physically obstructing vessels; classify every vessel as fast/slow/no-flow; and count adhered RBCs. Pre-transfusion SCD patients had 49% fast-flow vs 78% in controls (P = 5.8e-4), 16% no-flow vs 2% (P = 0.001), and 1.37 vs 0.01 adhered RBCs per vessel (P = 0.0025). Post-transfusion, fast-flow rose to 66% (P = 0.0098), no-flow fell to 6% (P = 0.0039), and adhered cells dropped by ~half (P = 0.043). Blackbird angle: (a) first quantitative, mechanistic pharmacodynamic biomarker for vaso-occlusion in a $4-5B SCD market where every Phase 2/3 program (Casgevy, Lyfgenia, crizanlizumab, voxelotor) suffers from noisy patient-reported VOC primary endpoints; (b) defensible IP is the instrument geometry plus the deep-learning analytic stack — closer to a Cytalux/Spectral MD device playbook than a small molecule; (c) clear pharma deal pathway — embed OBM as PD readout in pivotal sickle-cell trials at Vertex/CRISPR, Bluebird, Pfizer (Global Blood Therapeutics), Novartis; (d) right team — Durr ships clinical-grade imaging hardware, Pecker runs the Hopkins adult sickle program, Lanzkron now at Jefferson; (e) platform-extension: same OBM hardware reads out on any microvascular adhesion/flow disturbance — diabetic microvasculature, cerebral small-vessel disease, sepsis, CAR-T CRS. Open items: 10-patient cohort needs multi-site reproducibility; competitive landscape includes Sight Diagnostics' microfluidic label-free hematology (different form factor); Durr has not yet been a founding CEO. Recommended action: JHTV-routed meeting with Durr, Pecker, and TTREC neighbors on IP breadth, any active pharma sponsored-research, and an 18-month multi-site validation plan; then a Blackbird translational grant funding the validation, with BioHub as the incubation home for the spinout. Watch-list: Rossi, Sportelli, Kikidis, Hyde, Kleinman, Weinberger, Pergola (Lieber Institute + Hopkins) Nature Genetics paper (DOI 10.1038/s41588-026-02646-3; PMID 42332269; published June 22, 2026) developing INGENE and MODULE — two new co-expression-based TWAS models capturing distal trans-acting variant effects using the Lieber BrainSeq RNA-seq atlas (6 brain regions). Improves expression imputation for 18,744 genes; applied to PGC wave 3, calls 766 schizophrenia genes including 641 novel TWAS hits. Lieber platform-leverage story directly adjacent to the GPR52 NewCo thesis — ask Pergola/Weinberger/Hyde for the rank-ordered druggable-target shortlist (small-molecule-tractable GPCRs, ion channels, kinases) not already worked by BMS, AbbVie, Boehringer; and whether the INGENE-MODULE method itself is licensable as a software platform or as the analytical engine inside a Lieber-anchored target-discovery vehicle.

Sourcing Radar — Wilmer crystallizes a tyrosine kinase inhibitor into a long-acting wet AMD shot

Mon, 22 Jun 2026 12:00:00 +0000

Daily Sourcing Radar for June 22, 2026. Lead pick is a Wilmer Eye Institute paper in Journal of Controlled Release (Lu, Rakoski, Ri et al., corresponding Elia J. Duh and Joshua C. Doloff; DOI 10.1016/j.jconrel.2026.115120; PMID 42314988; published June 18, 2026). The team developed a crystalline solid-state formulation of sorafenib — the off-patent multi-receptor tyrosine kinase inhibitor (VEGFR1/2/3, PDGFR, RAF) — for intravitreal delivery. A single injection in two mouse models of neovascular AMD (LI-CNV and Vldlr-/-) suppressed both choroidal neovascularization and vascular leakage for at least 28 days, with ~13% of dose remaining in the eye at day 28 and no measurable adverse effect on intraocular pressure, OCT retinal morphology, or ERG. Blackbird angle: (a) wet AMD durability is the largest unmet need in a ~$13B market dominated by monthly aflibercept/ranibizumab/faricimab — Susvimo and Eylea HD have moved the bar partially, but room for a true 6-month injectable remains wide open; (b) sorafenib's multi-RTK profile is differentiated from anti-VEGF biologics and offers a clear wedge in the ~30% incomplete-responder population; (c) the off-patent molecule is de-risked but the crystal-form and formulation IP are defensible — the same playbook EyePoint Pharmaceuticals ran with vorolanib in DURAVYU, a public comp ~$1.5B on Phase 3 durability data; (d) senior team — Duh runs a serious Wilmer translational neovascular AMD program; Doloff and Scott Wilson are in Hopkins' Translational Therapeutics & Regenerative Engineering Center (TTREC, the Wilmer-BME joint operation), Doloff trained in the Langer/Anderson labs at MIT; (e) the authors explicitly position crystallization as a proof-of-concept platform for any hydrophobic ocular drug — a pipeline-in-a-formulation claim covering diabetic retinopathy, geographic atrophy, dry eye programs gated by solubility. Open items: mouse data only — need 6-month NHP/rabbit durability before sizing; crystal-depot inflammation/dispersion is the standard intraocular crystal risk; competitive set (DURAVYU, gene-therapy plays) is crowded. Recommended action: a JHTV-routed meeting with Duh, Doloff, and Wilson about IP breadth (single asset vs. platform claims), existing sponsored research, and the rabbit/NHP plan — then a translational grant into 6-month durability in large-animal models. Watch-list: Breen, Tao, Hyde, Kleinman et al. (Mount Sinai + Lieber Institute) Cell Reports DOI 10.1016/j.celrep.2026.117514 (PMID 42295976), published June 15 — large postmortem transcriptomics across 169 donors (86 control, 57 ED, 27 OCD), 0.67 (DLPFC) / 0.75 (caudate) transcriptome-wide correlation between ED and OCD; joint analysis nominates GABAergic interneurons (SST, parvalbumin), neuroendocrine (VGF, CRH), mitochondrial translation, and CHD8-interacting networks. Validates Lieber's postmortem-brain platform; anorexia has no approved pharmacology and highest psychiatric mortality. Check in with Tom Hyde / Dan Weinberger on whether anything chemistry-enabled is downstream.

Portfolio Watch — Hopkins opens an $80M-a-year science fund; GSK pays $10.6B for novel-target oncology

Sun, 21 Jun 2026 13:00:00 +0000

Weekly Portfolio Watch for the week ending June 21, 2026. Five items, three of which move the thesis. (1) Johns Hopkins announced its Life Sciences Research Initiative on June 10, 2026 — $80M/year over two years, with two tracks (Transformational Science Team Awards starting at $10M per project across four years; High-Impact Individual Awards) and a faculty oversight committee chaired by Ashani Weeraratna. The committee opened applications June 15. Implication: tailwind. This is the kind of multi-PI, milestone-driven funding that matures preclinical packages to the point where Blackbird Laboratories can grant in and BioVentures can write seed checks. The team should figure out how to get visibility on the submitted-project longlist. (2) GSK to acquire Nuvalent for $10.6B in cash, announced June 9, 2026, at $124/share — a 40% premium. Nuvalent is the precision-oncology shop behind next-generation ROS1 and ALK kinase inhibitors for non-small-cell lung cancer (zidesamtinib, September PDUFA; neladalkib, November PDUFA). Implications: validation comp for novel-target oncology assets (relevant to the Sidney Kimmel pipeline, including the DCTPP1 work on today's Sourcing Radar), and a competitive watch-item for aSKY's NSCLC positioning since both eventually target the same patient flow. (3) Neumora navacaprant — its kappa-opioid antagonist — failed both KOASTAL-2 and KOASTAL-3 Phase 3 trials in major depressive disorder, reported June 15, 2026. Neumora is killing the depression program, cutting ~35% of staff, and refocusing on Alzheimer's agitation, schizophrenia, and obesity. Net for the Lieber/Third Rock GPR52 schizophrenia NewCo: macro tailwind (the field is converging on non-dopaminergic mechanisms with all-three-symptom-domain ambition — exactly the GPR52 pitch), but Neumora becomes a louder schizophrenia competitor and is incentivized to partner aggressively on whatever asset they have left in the indication. (4) Lilly acquires 4E Therapeutics, June 16, 2026 — oral MNK/eIF4E pathway inhibitors for peripherally restricted non-opioid pain. Lead 4ET1103 finished Phase 1 with clean safety. Off-thesis directly, but continues Lilly's aggressive mechanism-first early-stage M&A pattern (Akouos, Seamless, now 4E) — the macro deal pattern Blackbird-incubated companies will eventually feed. (5) Incyte acquires Vega Therapeutics — Star Therapeutics subsidiary — for $1.25B up front, up to $2B total with milestones, announced June 8, 2026. Vega's VGA039 is a Phase 3 subcutaneous monoclonal for von Willebrand disease. Off-thesis but a reminder that hematology Phase 3 assets still price north of a billion. Sources: hub.jhu.edu/2026/06/10/life-sciences-research-initiative/; hub.jhu.edu/2026/06/15/research-funding-opportunities-application-period/; gsk.com/media/nehelih4/nuvalent-ir-call-slides-final-9-june-2026.pdf; biospace.com (Neumora navacaprant Phase 3 failure, June 15, 2026); 4E Therapeutics / PR Newswire June 16, 2026; Business Wire / Incyte June 8, 2026.

Sourcing Radar — Hopkins drug discovery cracks DCTPP1, unlocking decitabine in prostate cancer

Sun, 21 Jun 2026 12:00:00 +0000

Daily Sourcing Radar for June 21, 2026. Lead pick is a new PNAS paper from the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins — Hauk, Liu, Nelson, Yegnasubramanian, Berger; DOI 10.1073/pnas.2534029123; PMID 42296362; published June 15, 2026. The target is DCTPP1, a nucleotide pyrophosphatase that chews up the activated form of decitabine and accounts for cell-intrinsic resistance to nucleoside DNA methyltransferase inhibitors in solid tumors. The team had previously shown the biology with triptolide (bioRxiv 10.1101/2024.05.19.594134, 2024); the PNAS paper extends that work with (a) a high-throughput chemical screen yielding previously unreported inhibitor classes with submicromolar potency, (b) X-ray crystal structures of leads in the nucleotide-binding pocket mapping interactions with a tryptophan pair and two histidines, and (c) confirmed synergy with decitabine in prostate cancer cells. Blackbird angle: novel target with no clinical competition; defensible chemistry anchored by the crystal structures; commercial framing as a combination program that could broaden a generic decitabine (FDA-approved in MDS only) into castrate-resistant prostate cancer and other epigenetically-resistant tumors with high DCTPP1; senior team (Nelson, Yegnasubramanian, Berger) is the right team to drive a program. Open items: cell-based only in this paper, submicromolar not picomolar, and the natural-product reputation of the triptolide precursor sets a high cleanliness bar for the new scaffolds — but the structural-biology hook de-risks much of that. Recommended action: take a meeting with the Yegnasubramanian and Berger labs about a translational grant for in-vivo prostate xenografts and lead-series medchem optimization. Watch-list pick: a June 18 PNAS paper on endothelial KLF4 (Dhar et al., corresponding Pieper at Case Western, with Bindu D. Paul at Hopkins / Lieber Institute as a co-author; DOI 10.1073/pnas.2426990123; PMID 42313933) frames endothelial KLF4 as a lever on neurovascular decline and neuropsychiatric phenotypes — target ID, not a chemical lead, but worth tracking for the Lieber thesis adjacency.