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.

digital-trial/main_conplex.nf

@@ -0,0 +1,123 @@
+nextflow.enable.dsl=2
+
+
+process CONPLEX {
+    container "${params.container_conplex}"
+    containerOptions "${params.containerOptions}"
+    publishDir "${params.outdir}/${params.project_name}/receptors", mode: 'copy'
+    debug true
+    maxForks 5
+    // scratch true //deletes workdir after successful completion
+
+    input:
+     path smi_drug
+     path smi_metabolite
+     path mut_fasta
+
+    output:
+     path "*drug_scores.tsv", emit: drug
+     path "*significant_interactions.tsv", emit: interactions
+     
+    script:
+    """
+    workdir=`pwd`
+
+    #concat mutated and reference fasta files. If it has name with _2 its mutated.
+    #after test is don retrun this
+    cat $mut_fasta /home/omic/ConPLex/MANE_referent_transcipt_reference.fasta > \$workdir/all_protein.fasta
+
+    ### Test
+    #head -n 100 /home/omic/ConPLex/MANE_referent_transcipt_reference.fasta > \$workdir/test.fasta
+    #cat $mut_fasta \$workdir/test.fasta > \$workdir/all_protein.fasta
+    ###
+
+    . activate conplex-dti
+
+     #transform smi for conplex
+    python3 -c "
+import sys
+from rdkit import Chem
+import numpy as np
+import pandas as pd
+
+def converter(drug_csv, outname):
+    #open file to write drug/matabolite smi 
+    out_file = open(outname + '.txt', 'w')
+    
+    #load drug csv; only has drug in separate lines
+    drug_smi_list = np.array(pd.read_csv(drug_csv)['SMILES']).tolist()
+    #original drugs are anotated with drug_{}
+    for n, smi in enumerate(drug_smi_list):
+        out_file.write('{}\\tdrug_{}\\n'.format(smi, n))
+ 
+    out_file.close()
+
+converter('$smi_drug', 'drag')
+    "
+    #concatinate drag and metabolites
+    cat drag.txt $smi_metabolite > \$workdir/smiles.smi
+    #conplex predicition 1. round 
+    python3 /home/omic/ConPLex/conplex.py --fastas \$workdir/all_protein.fasta --smis \$workdir/smiles.smi --screening_batch_size ${params.screening_batch_size} --outdir \$workdir --split_write_df 'no'
+    #### this part is use to test all alternative transcripts (this part can be skipped for now, cuz we will use it for tissue specific analysis later) 
+    ##get fasta for 2. round transcripts for proteins above threshold
+    #python3 /home/omic/ConPLex/get_round_2.py --threshold ${params.threshold} --workdir \$workdir --round 1 
+    ##conplex all transcipt for protrin above threshold
+    #python3 /home/omic/ConPLex/conplex.py --fastas \$workdir/round_2.fasta --smis \$workdir/smiles.smi --screening_batch_size ${params.screening_batch_size} --outdir \$workdir --split_write_df 'no'
+    ####
+    #get data for second run
+    python3 /home/omic/ConPLex/get_round_2.py --threshold ${params.threshold} --workdir \$workdir --round 2  
+    """
+}
+
+process NETWORK_ENRICHMENT {
+  container "${params.container_conplex}"
+  containerOptions "${params.containerOptions}"
+  publishDir "${params.outdir}/${params.project_name}/receptors", mode: 'copy'
+  debug true
+  maxForks 5
+
+  input:
+    path interactions
+
+  output:
+    path "*_network_enrichment.tsv", emit: network_enrichment
+    path "*_network_interactions.tsv", emit: network_interactions
+
+  script:
+    """
+#!/opt/conda/envs/conplex-dti/bin/python 
+import pandas as pd
+import requests
+
+interaction_data = pd.read_csv('$interactions', sep = '\\t')
+interaction_data =  interaction_data.sort_values('conplex_score', ascending=False)
+
+#get list of enst above threshold
+protein_network_list = '%0d'.join(list(pd.unique(interaction_data[interaction_data['conplex_score'] > ${params.protein_network_threshold}]['transcipt'])))
+
+#get urls
+url_network = f'https://string-db.org/api/tsv/network?identifiers={protein_network_list}&species=9606'
+url_enrichment = f'https://string-db.org/api/tsv/enrichment?identifiers={protein_network_list}&species=9606'
+
+#get enrichment dataframe
+r = requests.get(url_enrichment)
+lines = r.text.split('\\n') # pull the text from the response object and split based on new lines
+data = [l.split('\\t') for l in lines] # split each line into its components based on tabs
+enrichment_df = pd.DataFrame(data[1:-1], columns = data[0]) # convert to dataframe using the first row as the column names; drop empty, final row
+#get protein interaction dataframe
+r = requests.get(url_network)
+lines = r.text.split('\\n') # pull the text from the response object and split based on new lines
+data = [l.split('\\t') for l in lines] # split each line into its components based on tabs
+# convert to dataframe using the first row as the column names; drop empty, final row
+interactions_df = pd.DataFrame(data[1:-1], columns = data[0]) 
+# dataframe with the preferred names of the two proteins and the score of the interaction
+#interactions_df = interactions_df[['preferredName_A', 'preferredName_B', 'score']]  
+
+#output name
+output_name = '$interactions'.split('/')[-1].split('_significant_interactions.tsv')[0]
+
+#save
+enrichment_df.to_csv(f'{output_name }_network_enrichment.tsv',sep='\\t')
+interactions_df.to_csv(f'{output_name }_network_interactions.tsv',sep='\\t')
+    """
+}