improvements to tethering. rmsd filter.

Author: tdudgeon

.dockerignore

@@ -4,3 +4,6 @@ gradle
 openshift
 work
 .nextflow
+trace.txt*
+report.html*
+wip

src/nextflow/xchem/prepare-tether-featurestein.nf

@@ -1,6 +1,7 @@
 #!/usr/bin/env nextflow
 
-params.candidates = "data/mpro/expanded-17.json"
+params.smiles = '*.smi'
+params.molfiles = '*.mol'
 params.fragments = "data/mpro/hits-17.sdf.gz"
 params.chunk_tether = 250
 params.chunk_score = 10000
@@ -9,7 +10,9 @@ params.digits = 4
 params.generate_filenames = false
 params.num_conformers = 10
 
-candidates = file(params.candidates)
+// files
+smilesfiles = file(params.smiles)
+molfiles = file(params.molfiles)
 fragments = file(params.fragments)
 
 process generate_feat_maps {
@@ -27,19 +30,27 @@ process generate_feat_maps {
     """
 }
 
-process split_json {
+process splitter {
 
     container 'informaticsmatters/rdkit_pipelines:latest'
 
     input:
-    file candidates
+    file smiles from smilesfiles.flatten()
+    file mol from molfiles.flatten()
 
     output:
-    file '*.smi' into smiles
     file '*.mol' into mols
+    file '*.smi' into smiles
 
     """
-    python -m pipelines.xchem.split_fragnet_candidates -i '$candidates' ${params.generate_filenames ? '--generate-filenames' : ''}
+    stem=${smiles.name[0..-5]}
+    split -l $params.chunk_tether -d -a 3 --additional-suffix .smi $smiles \${stem}_
+    mv $smiles ${smiles}.orig
+    for f in *.smi
+    do
+      cp $mol \${f:0:-4}.mol
+    done
+    mv $mol ${mol}.orig
     """
 }
 

src/nextflow/xchem/prepare-tether.nf

@@ -1,40 +1,63 @@
 #!/usr/bin/env nextflow
 
-params.candidates = "data/mpro/expanded-17.json"
-params.chunk = 1000
+params.smiles = '*.smi'
+params.molfiles = '*.mol'
+params.chunk_tether = 250
+params.chunk_score = 10000
 params.limit = 0
-params.digits = 4
-params.generate_filenames = false
+params.num_conformers = 10
+params.atom_compare = 'CompareElements'
+params.bond_compare = 'CompareOrder'
+params.complete_rings_only = true
+params.ring_matches_ring_only = true
+params.minimize = 4
 
-candidates = file(params.candidates)
 
-process split_json {
+smilesfiles = file(params.smiles)
+molfiles = file(params.molfiles)
+
+process splitter {
 
     container 'informaticsmatters/rdkit_pipelines:latest'
 
     input:
-    file candidates
+    file smiles from smilesfiles.flatten()
+    file mol from molfiles.flatten()
 
     output:
-    file '*.smi' into smiles
     file '*.mol' into mols
+    file '*.smi' into smiles
 
     """
-    python -m pipelines.xchem.split_fragnet_candidates -i '$candidates' ${params.generate_filenames ? '--generate-filenames' : ''}
+    stem=${smiles.name[0..-5]}
+    split -l $params.chunk_tether -d -a 3 --additional-suffix .smi $smiles \${stem}_
+    mv $smiles ${smiles}.orig
+    for f in *.smi
+    do
+      cp $mol \${f:0:-4}.mol
+    done
+    mv $mol ${mol}.orig
     """
 }
 
 process tether {
 
     container 'informaticsmatters/rdkit_pipelines:latest'
+    publishDir '.'
 
     input:
-    file smiles from smiles.flatten() //collect().toSortedList().flatten()
-    file mol from mols.flatten() //collect().toSortedList().flatten()
+    file smiles from smiles.flatten()
+    file mol from mols.flatten()
 
     output:
-    file 'Tethered_*.sdf'
+    file 'Tethered_*.sdf' into tethered_parts
+
     """
-    python -m pipelines.xchem.prepare_tether --smi '$smiles' --mol '$mol' -o 'Tethered_${smiles.name[0..-5]}'
+    python -m pipelines.xchem.prepare_tether --smi '$smiles' --mol '$mol' --chunk-size $params.chunk_score\
+      --num-conformers $params.num_conformers -o 'Tethered_${smiles.name[0..-5]}'\
+      --atom-compare $params.atom_compare --bond-compare $params.bond_compare\
+      ${params.complete_rings_only ? '--complete-rings-only' : ''}\
+      ${params.ring_matches_ring_only ? '--ring-matches-ring-only' : ''}\
+      --minimize $params.minimize
     """
 }