# Input Settings: initialize base input folder names,
# dataset collections, and algorithms to run over
input_settings:

    # Base input directory
    input_dir : "inputs/example"

    # datasets: A list of datasets. Each dataset has the following parameters:
    #   dataset_id: Identifier for the dataset group. Used to resolve the input path.
    #   should_run: whether or not to run this dataset.
    #   groundTruthNetwork: Name of the reference network file in edge list format.
    #       Shared across all runs within the dataset. Needed for evaluation.
    #   nickname: (optional) Short display label used by the plotter for plot
    #       titles and heatmap column headers. Defaults to dataset_id when absent.
    #   scan_run_subdirectories: (optional) If true, the pipeline automatically
    #       discovers runs by scanning all subdirectories of the dataset's input
    #       directory. Mutually exclusive with 'runs'; an error is raised if no
    #       subdirectories are found.
    #
    #   runs: A list of runs within the dataset, each with the following parameters:
    #       run_id: Identifier for this specific run. Used as the input subdirectory name.
    #       exprData: scRNA-Seq expression data file. Cells are along the
    #           columns and genes are along the rows.
    #       pseudoTimeData: a file containing pseudotime ordering, or any other
    #           information about cells.

    # Denotes a list of algorithms to run. Each has the following parameters:
    #   algorithm_id: Identifier for the algorithm. Must be recognized by the pipeline,
    #       see code for acceptable values.
    #
    #   should_run: whether or not to run the algorithm
    #
    #   image: Docker image name to run for this algorithm (e.g. "grnbeeline/genie3:base").
    #       Mandatory. See README for the default image name for each algorithm.
    #
    #   params: any additional, algorithm-specific parameters
    #       should be specified in the params map for a given algorithm
    #
    algorithms:
              

        - algorithm_id: "PIDC"
          image: "grnbeeline/pidc:base"
          should_run: [True]


        - algorithm_id: "GRNVBEM"
          image: "grnbeeline/grnvbem:base"
          should_run: [True]


        - algorithm_id: "GENIE3"
          image: "grnbeeline/arboreto:base"
          should_run: [True]


        - algorithm_id: "GRNBOOST2"
          image: "grnbeeline/arboreto:base"
          should_run: [True]


        - algorithm_id: "PPCOR"
          image: "grnbeeline/ppcor:base"
          should_run: [True]
          params:
              # p-value cutoff
              # Used in parsing output
              pVal: [0.01]


        - algorithm_id: "SCODE"
          image: "grnbeeline/scode:base"
          should_run: [True]
          params:
              z: [10]
              nIter: [1000]
              nRep: [6]


        - algorithm_id: "SINCERITIES"
          image: "grnbeeline/sincerities:base"
          should_run: [True]
          params:
              nBins: [10]


        - algorithm_id: "LEAP"
          image: "grnbeeline/leap:base"
          should_run: [True]
          params:
              # Default maxLag value is 0.33
              maxLag: [0.33]


        - algorithm_id: "GRISLI"
          image: "grnbeeline/grisli:base"
          should_run: [True]
          params:
              L: [10]
              R: [3000]
              alphaMin: [0.0]


        - algorithm_id: "SINGE"
          image: "grnbeeline/singe:0.4.1"
          should_run: [True]
          params:
              lambda: [0.01]
              dT: [15]
              num_lags: [5]
              kernel_width: [0.5]
              prob_zero_removal: [0]
              prob_remove_samples: [0.0]
              family: ["gaussian"]
              num_replicates: [6]


        - algorithm_id: "SCRIBE"
          image: "grnbeeline/scribe:base"
          should_run: [True]
          params:
              ### required parameters
              # a list of delay values
              delay: ["5"]
              # any of 'RDI', 'uRDI', 'cRDI', or 'ucRDI'
              method: ['ucRDI']
              # lower detection limit (expression below this
              # will be treated as zero.
              lowerDetectionLimit: [0]
              # expressionFamily: for synthetic data use uninormal
              #  for mRNA count data use negbinomial.size()
              expressionFamily: ['uninormal']
              ### optional but recommended parameters
              # log transform expression values or not
              log: [False]
              # ignore pseudotime values (and use experimental
              # time points instead), recommended True for synthetic data
              # False for real mRNA data
              ignorePT: [True]


        - algorithm_id: "PEARSON"
          image: "local"
          should_run: [True]


    datasets:
        - dataset_id: "GSD"
          should_run: [True]
          scan_run_subdirectories: True
          groundTruthNetwork: "GroundTruthNetwork.csv"
          # nickname: "Gonadal Sex Determination"  # optional: overrides dataset_id in plot labels
          # runs:
          #   - run_id: "ex1"
          #     exprData: "ExpressionData.csv"
          #     pseudoTimeData: "PseudoTime.csv"

# Output Settings: configure where results are written
output_settings:

    # Base output directory
    output_dir: "outputs"
    # Optional: if set, outputs are written to output_dir/experiment_id/...
    # so multiple experiment runs can coexist under the same base directory.
    experiment_id: "example_run_GSD"