Merge pull request #6 from jjc2718/experiment_script

Add script to run pancancer classification experiments

.github/workflows/tests.yaml

@@ -1,6 +1,6 @@
 name: Python tests
 
-on: [push, pull_request]
+on: [pull_request]
 
 jobs:
     run-tests:

.gitignore

@@ -4,3 +4,6 @@
 
 data/EBPlusPlusAdjustPANCAN_IlluminaHiSeq_RNASeqV2-v2.geneExp.tsv
 data/*.gz
+data/*.pkl
+
+results/*

01_run_pancancer_classification.py

@@ -0,0 +1,49 @@
+"""
+Script to run pancancer classification experiments, by calling classify_cancer_type.py
+with the corresponding arguments
+"""
+import os
+import sys
+import subprocess
+
+import pandas as pd
+from tqdm import tqdm
+
+import pancancer_utilities.config as cfg
+
+EXP_SCRIPT = os.path.join(cfg.repo_root,
+                          'pancancer_utilities',
+                          'scripts',
+                          'classify_cancer_type.py')
+
+# genes and cancer types to run experiments for
+# just hardcoding these for now, might choose them systematically later
+genes = ['TP53', 'PTEN', 'KRAS', 'BRAF', 'TTN']
+cancer_types = ['BRCA', 'THCA', 'SKCM', 'GBM', 'SARC']
+
+def run_single_experiment(gene, cancer_type, use_pancancer,
+                          shuffle_labels, verbose=False):
+    args = [
+        'python',
+        EXP_SCRIPT,
+        '--gene', gene,
+        '--holdout_cancer_type', cancer_type
+    ]
+    if use_pancancer:
+        args.append('--use_pancancer')
+    if shuffle_labels:
+        args.append('--shuffle_labels')
+
+    if verbose:
+        print('Running: {}'.format(' '.join(args)))
+    subprocess.call(args)
+
+if __name__ == '__main__':
+
+    for gene in tqdm(genes):
+        for cancer_type in tqdm(cancer_types):
+           run_single_experiment(gene, cancer_type, False, False)
+           run_single_experiment(gene, cancer_type, True, False)
+           run_single_experiment(gene, cancer_type, False, True)
+           run_single_experiment(gene, cancer_type, True, True)
+