Digital Trials pipeline configured for WES

Source-only snapshot of the cluster branch for WES execution. Large reference files (HPA/MANE/ensemble FASTA, model weights, ~597 MB) are omitted: they are baked into the container images at build time and mounted from the dreamdock-data PVC at runtime, and exceed the Gitea request size limit. Pipeline entry point is main.nf, which orchestrates the biotransformer, conplex and tissue modules as a single workflow. Ligand inputs are read from the eureka workspace; protein_zarr and chembl_db come from the dreamdock-data PVC.
2026-07-27 21:59:52 +01:00
commit 9e75f44f1a
86 changed files with 10142 additions and 0 deletions
--- a/drug_tissue_distribution.py
+++ b/drug_tissue_distribution.py
@@ -0,0 +1,67 @@
+from pathlib import Path
+import argparse 
+import pandas as pd
+import numpy as np
+
+def get_tissue_distribution(file_name):
+    # Convert to Path object
+    file_path = Path(file_name)
+    
+    # Load interaction data — only the 'transcipt' column is referenced downstream
+    # (the bulk of the file is the 'smile' column we never use). Skipping it cuts
+    # memory ~30x on metabolite-heavy drugs (11 GB file -> ~400 MB resident).
+    significant_interactions = pd.read_csv(file_path, sep='\t', usecols=['transcipt'])
+    
+    # Check if data exists
+    if len(significant_interactions) == 0:
+        print(f"No significant interactions found in {file_name}. Creating empty output.")
+        tissue_per_prot_symbol_pd = pd.DataFrame()
+        save_name = str(file_path.parent / (file_path.stem.replace('_significant_interactions', '_tissue_distribution') + '.tsv'))
+        tissue_per_prot_symbol_pd.to_csv(save_name, sep='\t')
+        print(f"Saved empty tissue distribution to: {save_name}")
+        return
+    
+    # Load protein expression per tissue
+    HPA = pd.read_csv('/home/omic/HPA_normal_ihc_data.tsv', sep='\t')
+    # Load enst symbol mapping data
+    MANE_all_transcipts = pd.read_csv('/home/omic/MANE_all_transcipts.csv')
+
+    # Filter out tissues with low confidence
+    HPA = HPA[(HPA['Level'] == 'Medium') | (HPA['Level'] == 'High')]
+
+
+    # Get protein symbols from significant interactions
+    prot_enst = [i.split('_')[0] for i in significant_interactions['transcipt'].unique()]
+    
+    # Safely map to symbols, skipping missing ones
+    prot_symbol = []
+    for i in prot_enst:
+        matches = MANE_all_transcipts[MANE_all_transcipts['transcipt'] == i]
+        if not matches.empty:
+            prot_symbol.append(matches['symbol'].iloc[0])
+        else:
+            print(f"Warning: Transcript {i} not found in MANE mapping. Skipping.")
+            
+    prot_symbol = pd.unique(prot_symbol)
+
+    # Get unique tissues
+    unique_tissues = HPA['Tissue'].unique()
+    
+    # Get protein expression per tissue
+    tissue_per_prot_symbol = [list(HPA[HPA['Gene name'] == i]['Tissue'].unique()) for i in prot_symbol]
+    # Transform list to array
+    tissue_per_prot_symbol_array = np.array([[tissue in gene_tissues for tissue in unique_tissues] for gene_tissues in tissue_per_prot_symbol])
+    
+    tissue_per_prot_symbol_pd = pd.DataFrame(tissue_per_prot_symbol_array, columns=unique_tissues, index=prot_symbol)
+    
+    # Save
+    save_name = str(file_path.parent / (file_path.stem.replace('_significant_interactions', '_tissue_distribution') + '.tsv'))
+    tissue_per_prot_symbol_pd.to_csv(save_name, sep='\t')
+    print(f"Saved tissue distribution to: {save_name}")
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Get tissue distribution from *significant_interactions.tsv.")
+    parser.add_argument("--file_name", required=True, help="file_name")
+    args = parser.parse_args()
+    
+    get_tissue_distribution(args.file_name)