context-enrichment.conseq

rule create_contexts: 
    inputs: 
        in={'type': 'cell_line_metadata'},
        make_context_matrix=fileref('scripts/make_context_matrix.py'),
        cleanup_script=fileref('scripts/cleanup_hdf5.py')
    outputs: {
      'type': 'context-matrix',
      'hdf5_filename': {"$filename": "contexts.hdf5"},
      'csv_filename': {"$filename": "out.csv"},
      'dataset_id': '{{inputs.in.sample_info_id}}', # this is named dataset_id, because this key is recognized by the db load
    },
    { 'type': 'biomarker-matrix',
      'source_dataset_id': '{{inputs.in.sample_info_id}}',
      'category': 'context',
      'filename': {'$filename': 'contexts.hdf5'}
    }
    run "python3 {{inputs.make_context_matrix.filename}} {{inputs.in.filename}} temp-out.hdf5 out.csv"
    run "python3 {{inputs.cleanup_script.filename}} temp-out.hdf5 contexts.hdf5 --add-arxspan-to-cols -m {{inputs.in.filename}}"


rule context_enrichment:
    executor: dsub 
    inputs: deps = all {'type': "dep-matrix", 'label' ~ '^(?!Repurposing_secondary_dose).*$'}, # negative lookahead to not run on secondary dose
        contexts = {"type": "context-matrix"}, 
        compute_enrichment=fileref('scripts/compute_enrichment.py', copy_to='compute_enrichment.py'),
        hdf5_utils=fileref('scripts/hdf5_utils.py', copy_to='hdf5_utils.py')
    outputs: {"type": "context-enrichment", 
              "filename": {"$filename": "t-test-enrichment.csv"}}
    run """python compute_enrichment.py \
      --dep_datasets "[{% for dep in inputs.deps %}
{\"filename\":\"{{dep.filename|quoted}}\",\"label\":\"{{dep.label|quoted}}\"}
{% if not loop.last %},{% endif %}
{% endfor %}]" \
      --context_matrix {{ inputs.contexts.hdf5_filename | quoted }}"""