generated: '2026-08-23'
method: searched
source: >-
  https://github.com/ksqtx/TRACE and its published command reference
  https://github.com/ksqtx/TRACE/blob/main/CLI_commands.md, read on 2026-08-23.
note: >-
  This is a first-party command-line tool published by KSQ Therapeutics in its own GitHub
  organization. It is NOT an API client — KSQ ships no API for a CLI to wrap. It is a bioinformatics
  analysis tool for single-cell RNA-seq classification, and it is recorded because it is real
  developer-facing software the company publishes, not because it exercises the content API
  catalogued elsewhere in this repo.
name: trace
binary: trace
repository: https://github.com/ksqtx/TRACE
docs: https://github.com/ksqtx/TRACE/blob/main/CLI_commands.md
language: python
framework: Click
version: 0.1.0
license: GPL-3.0 (repository) / MIT (pyproject.toml)
wraps_api: false
install:
- method: git
  command: pip install git+https://github.com/ksqtx/TRACE.git
  note: Not published to PyPI. requires-python >=3.11.
global_options:
- flag: --log-level
  values:
  - DEBUG
  - INFO
  - WARNING
  - ERROR
  - CRITICAL
  default: INFO
- flag: --config
  description: Path to a YAML configuration file.
- flag: --output-dir
  description: Output directory for results and artifacts.
  default: ./output
commands:
- group: meta
  commands:
  - name: version
    description: Display TRACEpy and Python version information.
  - name: init-config
    description: Generate a configuration template (default or batch).
  - name: validate-config
    description: Validate a configuration file's structure and values.
- group: data
  commands:
  - name: preprocess
    description: >-
      Preprocess single-cell RNA-seq data from h5ad files — gene selection plus clone-level or
      cell-level expression summarisation for downstream ML.
  - name: chunk-remote-h5ad
    description: Chunk a remote h5ad dataset for processing.
- group: model
  commands:
  - name: train
    description: Train a classifier on preprocessed expression data.
  - name: evaluate
    description: Evaluate a trained model.
  - name: compare
    description: Compare models.
  - name: retrain
    description: Retrain an existing model.
  - name: predict
    description: Run predictions with a trained model.
  - name: batch
    description: Run a batch workflow from a batch configuration.
- group: analysis
  commands:
  - name: analyze-gene-significance
    description: Analyse gene-level significance in a trained model.
  - name: classify-cells
    description: Classify cells from single-cell input.
command_count: 13