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.

nf_metabol_screen_adaptive/app/old_get_round_20.py

@@ -0,0 +1,143 @@
+import argparse
+import pandas as pd
+import numpy as np
+from itertools import chain
+from pathlib import PosixPath as Path
+
+seq_exist = pd.read_csv('/app/MANE_all_transcipts.csv')
+
+def get_round_2(threshold, workdir, round, drug_csv : Path):
+    workdir = Path(workdir)
+    
+    # Get output names
+    # Get patient fasta name
+    # fasta_files = list(workdir.glob('*_variants_transcript_id_mutations.fasta'))
+    # if not fasta_files:
+    #     raise FileNotFoundError("No variants transcript mutations fasta file found")
+    # fasta_name = fasta_files[0].stem.replace('_variants_transcript_id_mutations', '')
+    fasta_name = "patient_0"
+    
+    # Get test drug
+    drug_name = drug_csv.stem
+    
+    name_out = f"{drug_name}_{fasta_name}"
+    
+    # Get all work dir complex files
+    results_files = list(workdir.glob('*_results.tsv'))
+    if not results_files:
+        raise FileNotFoundError("No results TSV files found")
+    
+    round_1_score_list = [str(f) for f in results_files]
+    
+    # Read 1st round
+    transcipts_1 = [
+        pd.read_csv(i, sep='\t', header=None).sort_values([2], ascending=False)
+        for i in round_1_score_list
+        ]
+    
+    # Get position of drug interaction vs all metabolites
+    drug_pos = [
+        [
+            n
+            for n, j in enumerate(list(i[0]))
+            if j[:4] == 'drug'
+        ] 
+        for i in transcipts_1
+    ]
+    
+    # Drug score
+    drug_0_score = pd.concat([i[i[0] =='drug_0'] for i in transcipts_1]).rename({0:'Drug', 1:'Transcript', 2:'Score'}, axis=1)
+    
+    # Filter all below threshold
+    transcipts_1 = [i[i[2] > threshold] for i in transcipts_1]
+    
+    # Save transcript_complex above threshold to one file
+    if round == 2:
+        inter_import = pd.concat(transcipts_1, ignore_index=True).rename({0:'drug/metabolite',1:'transcipt',2:'conplex_score'},axis=1)
+        # Add drug
+        test_smiles = pd.read_csv(workdir / 'smiles.smi', sep='\t', header=None)
+        smi_ = [test_smiles[test_smiles[1] == i].iloc[0][0] for i in list(inter_import['drug/metabolite'])]
+        inter_import['smile'] = smi_
+        inter_import[['drug/metabolite','smile','transcipt','conplex_score']].to_csv(
+            workdir / f'{name_out}_significant_interactions.tsv', sep='\t', index=False
+        )
+        drug_0_score.to_csv(workdir / f'{name_out}_drug_scores.tsv', sep='\t', index=False)
+    
+    transcipts_1 = pd.DataFrame([
+        (Path(j).stem.replace('_results', ''), i.shape[0], i[2].mean(), k[0]) 
+        for i, j, k in zip(transcipts_1, round_1_score_list, drug_pos) 
+        if i.shape[0] != 0
+    ])
+    
+    # Check if any interaction is above threshold for first round 
+    if round == 1:
+        if transcipts_1.shape == (0,0):
+            transcipts_1.to_csv(workdir / 'round_1.csv', index=False)
+            with open(workdir / 'round_1.fasta', 'w') as f:
+                f.write("all_data_is_filtered_out\n")
+            return 'STOP NO TRANSCIPTS ABOVE THRESHOLD'
+
+    transcipts_1 = transcipts_1.rename({
+        0:'transcipt_name',
+        1:'number_of_iteracting_compounds', 
+        2:'mean_binding_above_threshold', 
+        3:'drug_position'
+    }, axis='columns')
+    
+    path_out = workdir / f'round_{round}.csv'
+    
+    if len(transcipts_1) == 0:
+        with path_out.open("w") as f:
+            f.write("NO TRANSCIPTS ABOVE THRESHOLD")
+            return 'STOP NO TRANSCIPTS ABOVE THRESHOLD'
+        
+    
+    transcipts_1['if_drug_above_threshold'] = transcipts_1.iloc[:,1] > transcipts_1.iloc[:,3]
+    
+    # If protein is mutated it has _2 in name, removes it
+    transcipt_names = [i.split('_')[0] for i in list(transcipts_1['transcipt_name'])]
+    temp = [
+        seq_exist[seq_exist['transcipt'] == i]
+        for i in transcipt_names
+    ]
+    transcipts_1['protein_name'] =[
+        i.iloc[0]['symbol'] if len(i) > 0 else ""
+        for i in temp
+    ]
+    
+    # Save data on first round
+    transcipts_1.to_csv(path_out, index=False)
+    
+    # Get fasta for second round
+    if round == 1:
+        # Get all transcripts of proteins above threshold
+        name_2_filttered = list(transcipts_1['protein_name'])
+        transcipts_2 = [list(seq_exist[seq_exist['symbol'] == i]['transcipt']) for i in name_2_filttered]
+        transcipts_2 = list(chain(*transcipts_2))
+        transcipts_2 = list(np.unique((transcipts_2)))
+        
+        # Filter out transcripts already ran through complex
+        transcipts_2 = list(np.array(transcipts_2)[
+            [i not in list(transcipts_1['transcipt_name']) for i in transcipts_2]
+        ])
+        
+        fasta_new = [
+            ['>'+i, seq_exist[seq_exist['transcipt'] == i]['seq'].iloc[0]] 
+            for i in transcipts_2
+        ]
+        fasta_new = list(chain(*fasta_new))
+        
+        # Write fasta to run
+        with open(workdir / 'round_2.fasta', 'w') as f:
+            for line in fasta_new:
+                f.write(f"{line}\n")
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Process and screen fragments.")
+    parser.add_argument("--threshold", required=True, type=float, help="Threshold for 1. round.of conplex scores")
+    parser.add_argument("--workdir", required=True, type=Path, help="workdir")
+    parser.add_argument("--round", required=True, type=int, help="Round 1 or 2")
+    parser.add_argument("--drug-csv", required=True, type=Path, help="csv file with drug smiles")
+    args = parser.parse_args()
+
+    get_round_2(args.threshold, args.workdir, args.round, args.drug_csv)