Functional Group Calculation (#44)

Co-authored-by: Tim Lorsbach <tim@lorsba.ch>
Reviewed-on: enviPath/enviPy#44

epdb/models.py

@@ -1438,6 +1438,8 @@ class ApplicabilityDomain(EnviPathModel):
     reliability_threshold = models.FloatField(blank=False, null=False, default=0.5)
     local_compatibilty_threshold = models.FloatField(blank=False, null=False, default=0.5)
 
+    functional_groups = models.JSONField(blank=True, null=True, default=dict)
+
     @staticmethod
     @transaction.atomic
     def create(mlrr: MLRelativeReasoning, num_neighbours: int = 5, reliability_threshold: float = 0.5,
@@ -1474,6 +1476,15 @@ class ApplicabilityDomain(EnviPathModel):
         ad = ApplicabilityDomainPCA(num_neighbours=self.num_neighbours)
         ad.build(ds)
 
+        # Collect functional Groups together with their counts for reactivity center highlighting
+        functional_groups_counts = defaultdict(int)
+        for cs in CompoundStructure.objects.filter(compound__package__in=self.model.data_packages.all()):
+            for fg in FormatConverter.get_functional_groups(cs.smiles):
+                functional_groups_counts[fg] += 1
+
+        self.functional_groups = dict(functional_groups_counts)
+        self.save()
+
         end = datetime.now()
         logger.debug(f"fitting app domain pca took {(end - start).total_seconds()} seconds")
 

utilities/chem.py

@@ -12,6 +12,7 @@ from rdkit.Chem import MACCSkeys
 from rdkit.Chem import rdChemReactions
 from rdkit.Chem.Draw import rdMolDraw2D
 from rdkit.Chem.MolStandardize import rdMolStandardize
+from rdkit.Contrib.IFG import ifg
 
 logger = logging.getLogger(__name__)
 RDLogger.DisableLog('rdApp.*')
@@ -87,6 +88,21 @@ class FormatConverter(object):
         bitvec = MACCSkeys.GenMACCSKeys(mol)
         return bitvec.ToList()
 
+    @staticmethod
+    def get_functional_groups(smiles: str) -> List[str]:
+        res = list()
+
+        try:
+            m = Chem.MolFromSmiles(smiles)
+            fgs = ifg.identify_functional_groups(m)
+            for fg in fgs:
+                # TODO atoms or type?
+                res.append(fg.atoms)
+        except AttributeError:
+            logger.debug(f"Could not get functional groups for {smiles}")
+
+        return res
+
     @staticmethod
     def to_svg(smiles, mol_size=(200, 150), kekulize=True):
         mol = FormatConverter.from_smiles(smiles)