add dude processing scripts

dude.py

@@ -0,0 +1,74 @@
+from mpi4py import MPI
+from mpi4py.futures import MPICommExecutor
+
+from Bio.PDB import PDBParser, PPBuilder
+import warnings
+
+import gzip
+import tempfile
+import os
+from rdkit import Chem
+
+import pandas as pd
+
+def open_ligands(fn):
+    with tempfile.NamedTemporaryFile(mode='w+b',delete=False) as f:
+        with gzip.open(fn,'rb') as g:
+            f.write(g.read())
+        name = f.name
+
+    suppl = Chem.SDMolSupplier(name)
+    os.unlink(name)
+    return suppl
+
+def get_ligands(path):
+    try:
+        parser = PDBParser()
+        with warnings.catch_warnings():
+            warnings.simplefilter("ignore")
+            structure = parser.get_structure('protein',path+'/receptor.pdb')
+        ppb = PPBuilder()
+        seq = []
+        for pp in ppb.build_peptides(structure):
+            seq.append(str(pp.get_sequence()))
+        seq = ''.join(seq)
+
+        name = os.path.basename(path)
+
+        decoys = open_ligands(path+'/decoys_final.sdf.gz')
+        actives = open_ligands(path+'/actives_final.sdf.gz')
+
+        actives_smiles = []
+        for m in actives:
+            try:
+                actives_smiles.append(Chem.MolToSmiles(m))
+            except:
+                 pass
+
+        decoys_smiles = []
+        for m in decoys:
+            try:
+                decoys_smiles.append(Chem.MolToSmiles(m))
+            except:
+                 pass
+
+        all_smiles = actives_smiles + decoys_smiles
+        all_active = [True]*len(actives_smiles) + [False]*len(decoys_smiles)
+        names = [name]*len(all_active)
+        seqs = [seq]*len(all_active)
+        df = pd.DataFrame({'name': names, 'seq': seqs, 'smiles': all_smiles, 'active': all_active})
+
+        df.to_parquet(path+'/ligands.parquet')
+    except Exception as e:
+        print(e)
+        pass
+
+
+if __name__ == '__main__':
+    import glob
+
+    filenames = glob.glob('DUDE/all/*')
+    comm = MPI.COMM_WORLD
+    with MPICommExecutor(comm, root=0) as executor:
+        if executor is not None:
+            executor.map(get_ligands, filenames)

dude.slurm

@@ -0,0 +1,9 @@
+#!/bin/bash
+#SBATCH -J preprocess_dude
+#SBATCH -p gpu
+#SBATCH -A STF006
+#SBATCH -t 3:00:00
+#SBATCH -N 8
+#SBATCH --ntasks-per-node=16
+
+srun python dude.py

dude_combine.py

@@ -0,0 +1,10 @@
+import dask.dataframe as dd
+
+if __name__ == '__main__':
+    import glob
+
+    filenames = glob.glob('DUDE/all/*/ligands.parquet')
+
+    ddf = dd.read_parquet(filenames)
+    df = ddf.compute()
+    df.to_parquet('data/dude.parquet')