feat: Add CIP annotations #136

app/modules/cdkmodules.py

@@ -18,18 +18,25 @@
 if not isJVMStarted():
     cdk_path = "https://github.com/cdk/cdk/releases/download/cdk-2.8/cdk-2.8.jar"
     sru_path = "https://github.com/JonasSchaub/SugarRemoval/releases/download/v1.3.2/SugarRemovalUtility-jar-with-dependencies.jar"
+    centres_path = (
+        "https://github.com/SiMolecule/centres/releases/download/1.0/centres.jar"
+    )
     cdkjar_path = str(pystow.join("STOUT-V2")) + "/cdk-2.8.jar"
     srujar_path = (
         str(pystow.join("STOUT-V2")) + "/SugarRemovalUtility-jar-with-dependencies.jar"
     )
+    centresjar_path = str(pystow.join("STOUT-V2")) + "/centres.jar"
 
     if not os.path.exists(cdkjar_path):
         jar_path = pystow.ensure("STOUT-V2", url=cdk_path)
 
     if not os.path.exists(srujar_path):
         jar_path = pystow.ensure("STOUT-V2", url=sru_path)
 
-    startJVM("-ea", classpath=[cdkjar_path, srujar_path])
+    if not os.path.exists(centresjar_path):
+        jar_path = pystow.ensure("STOUT-V2", url=centres_path)
+
+    startJVM("-ea", classpath=[cdkjar_path, srujar_path, centresjar_path])
     cdk_base = "org.openscience.cdk"
 
 
@@ -271,3 +278,101 @@ def getTanimotoSimilarityCDK(smiles1: str, smiles2: str):
     Similarity = Tanimoto.calculate(fingerprint1, fingerprint2)
 
     return "{:.5f}".format(float(str(Similarity)))
+
+
+def getCIPAnnotation(smiles: str):
+    """
+    The function return the CIP annotations using the CDK
+    CIP toolkit.
+    Args: mol block
+    Returns: CIP annotated mol block
+    """
+    mol = getCDKSDG(smiles)
+    centres_base = "com.simolecule.centres"
+    Cycles = JClass(cdk_base + ".graph.Cycles")
+    IBond = JClass(cdk_base + ".interfaces.IBond")
+    IStereoElement = JClass(cdk_base + ".interfaces.IStereoElement")
+    Stereocenters = JClass(cdk_base + ".stereo.Stereocenters")
+    StandardGenerator = JClass(
+        cdk_base + ".renderer.generators.standard.StandardGenerator"
+    )
+
+    BaseMol = JClass(centres_base + ".BaseMol")
+    CdkLabeller = JClass(centres_base + ".CdkLabeller")
+    Descriptor = JClass(centres_base + ".Descriptor")
+
+    stereocenters = Stereocenters.of(mol)
+    for atom in mol.atoms():
+        if (
+            stereocenters.isStereocenter(atom.getIndex())
+            and stereocenters.elementType(atom.getIndex())
+            == Stereocenters.Type.Tetracoordinate
+        ):
+            atom.setProperty(StandardGenerator.ANNOTATION_LABEL, "(?)")
+
+    for bond in mol.bonds():
+        if bond.getOrder() != IBond.Order.DOUBLE:
+            continue
+        begIdx = bond.getBegin().getIndex()
+        endIdx = bond.getEnd().getIndex()
+        if (
+            stereocenters.elementType(begIdx) == Stereocenters.Type.Tricoordinate
+            and stereocenters.elementType(endIdx) == Stereocenters.Type.Tricoordinate
+            and stereocenters.isStereocenter(begIdx)
+            and stereocenters.isStereocenter(endIdx)
+        ):
+            # only if not in a small ring <7
+            if Cycles.smallRingSize(bond, 7) == 0:
+                bond.setProperty(StandardGenerator.ANNOTATION_LABEL, "(?)")
+    # no defined stereo?
+    if not mol.stereoElements().iterator().hasNext():
+        return mol
+
+    CdkLabeller.label(mol)
+
+    # update to label appropriately for racmic and relative stereochemistry
+    for se in mol.stereoElements():
+        if se.getConfigClass() == IStereoElement.TH and se.getGroupInfo() != 0:
+            focus = se.getFocus()
+            label = focus.getProperty(BaseMol.CIP_LABEL_KEY)
+            if (
+                isinstance(label, Descriptor)
+                and label != Descriptor.ns
+                and label != Descriptor.Unknown
+            ):
+                if (se.getGroupInfo() & IStereoElement.GRP_RAC) != 0:
+                    inv = None
+                    if label == Descriptor.R:
+                        inv = Descriptor.S
+                    elif label == Descriptor.S:
+                        inv = Descriptor.R
+                    if inv is not None:
+                        focus.setProperty(
+                            BaseMol.CIP_LABEL_KEY, label.toString() + inv.name()
+                        )
+                elif (se.getGroupInfo() & IStereoElement.GRP_REL) != 0:
+                    if label in [Descriptor.R, Descriptor.S]:
+                        focus.setProperty(BaseMol.CIP_LABEL_KEY, label.toString() + "*")
+
+    for atom in mol.atoms():
+        if atom.getProperty(BaseMol.CONF_INDEX) is not None:
+            atom.setProperty(
+                StandardGenerator.ANNOTATION_LABEL,
+                StandardGenerator.ITALIC_DISPLAY_PREFIX
+                + str(atom.getProperty(BaseMol.CONF_INDEX)),
+            )
+        elif atom.getProperty(BaseMol.CIP_LABEL_KEY) is not None:
+            atom.setProperty(
+                StandardGenerator.ANNOTATION_LABEL,
+                StandardGenerator.ITALIC_DISPLAY_PREFIX
+                + str(atom.getProperty(BaseMol.CIP_LABEL_KEY)),
+            )
+    for bond in mol.bonds():
+        if bond.getProperty(BaseMol.CIP_LABEL_KEY) is not None:
+            bond.setProperty(
+                StandardGenerator.ANNOTATION_LABEL,
+                StandardGenerator.ITALIC_DISPLAY_PREFIX
+                + bond.getProperty(BaseMol.CIP_LABEL_KEY),
+            )
+
+    return mol

app/modules/depict.py

@@ -2,11 +2,13 @@
 from rdkit import Chem
 from rdkit.Chem import rdDepictor
 from rdkit.Chem.Draw import rdMolDraw2D
-from app.modules.cdkmodules import getCDKSDG
+from app.modules.cdkmodules import getCDKSDG, getCIPAnnotation
 from jpype import JClass
 
 
-def getCDKDepiction(smiles: str, molSize=(512, 512), rotate=0, unicolor=False):
+def getCDKDepiction(
+    smiles: str, molSize=(512, 512), rotate=0, CIP=True, unicolor=False
+):
     """This function takes the user input SMILES and Depicts it
        using the CDK Depiction Generator.
     Args:
@@ -25,7 +27,6 @@ def getCDKDepiction(smiles: str, molSize=(512, 512), rotate=0, unicolor=False):
         DepictionGenerator = (
             JClass(cdk_base + ".depict.DepictionGenerator")()
             .withSize(molSize[0], molSize[1])
-            .withAtomValues()
             .withParam(StandardGenerator.StrokeRatio.class_, 1.0)
             .withAnnotationColor(Color.BLACK)
             .withParam(StandardGenerator.AtomColor.class_, UniColor(Color.BLACK))
@@ -37,15 +38,17 @@ def getCDKDepiction(smiles: str, molSize=(512, 512), rotate=0, unicolor=False):
             JClass(cdk_base + ".depict.DepictionGenerator")()
             .withAtomColors(CDK2DAtomColors)
             .withSize(molSize[0], molSize[1])
-            .withAtomValues()
             .withParam(StandardGenerator.StrokeRatio.class_, 1.0)
             .withFillToFit()
             .withBackgroundColor(Color.WHITE)
         )
     if any(char.isspace() for char in smiles):
         smiles = smiles.replace(" ", "+")
 
-    moleculeSDG = getCDKSDG(smiles)
+    if CIP:
+        moleculeSDG = getCIPAnnotation(smiles)
+    else:
+        moleculeSDG = getCDKSDG(smiles)
 
     if moleculeSDG:
 

app/routers/chem.py

@@ -150,10 +150,7 @@ async def CDK2D_Coordinates(smiles: str):
     if smiles:
         mol = Chem.MolFromSmiles(smiles)
         if mol:
-            return Response(
-                content=getCDKSDGMol(smiles).replace("$$$$\n", ""),
-                media_type="text/plain",
-            )
+            return getCDKSDGMol(smiles)
         else:
             return "Error reading SMILES string, check again."
 
@@ -193,6 +190,8 @@ async def Depict2D_molecule(
     width: Optional[int] = 512,
     height: Optional[int] = 512,
     rotate: Optional[int] = 0,
+    CIP: Optional[bool] = True,
+    unicolor: Optional[bool] = False,
 ):
     """
     Generate 2D Depictions using CDK or RDKit using given parameters.
@@ -206,7 +205,7 @@ async def Depict2D_molecule(
     if generator:
         if generator == "cdksdg":
             return Response(
-                content=getCDKDepiction(smiles, [width, height], rotate),
+                content=getCDKDepiction(smiles, [width, height], rotate, CIP, unicolor),
                 media_type="image/svg+xml",
             )
         else: