Add property calculator, molecular clustering, and scaffold analysis features

app.py

@@ -737,6 +737,138 @@ def name_to_3d_molecule(name: str) -> str:
         raise gr.Error(f"Error creating 3D molecule: {str(e)}")
 
 
+def calculate_properties_batch(smiles_list: str) -> str:
+    """Calculate physicochemical properties for multiple molecules"""
+    from rdkit.Chem import Lipinski
+    
+    lines = [line.strip() for line in smiles_list.strip().split('\n') if line.strip()]
+    if not lines:
+        return "Please enter at least one SMILES string (one per line)"
+    
+    results = []
+    results.append("SMILES\tMW\tLogP\tTPSA\tHBD\tHBA\tRotBonds\tRings\tAromRings")
+    
+    for smiles in lines[:50]:  # Limit to 50 molecules
+        try:
+            mol = Chem.MolFromSmiles(smiles)
+            if mol is None:
+                results.append(f"{smiles}\tInvalid SMILES")
+                continue
+            
+            mw = Descriptors.MolWt(mol)
+            logp = Descriptors.MolLogP(mol)
+            tpsa = Descriptors.TPSA(mol)
+            hbd = Lipinski.NumHDonors(mol)
+            hba = Lipinski.NumHAcceptors(mol)
+            rotbonds = Lipinski.NumRotatableBonds(mol)
+            rings = Lipinski.RingCount(mol)
+            arom_rings = Lipinski.NumAromaticRings(mol)
+            
+            results.append(f"{smiles}\t{mw:.2f}\t{logp:.2f}\t{tpsa:.2f}\t{hbd}\t{hba}\t{rotbonds}\t{rings}\t{arom_rings}")
+        except Exception as e:
+            results.append(f"{smiles}\tError: {str(e)}")
+    
+    return "\n".join(results)
+
+
+def cluster_molecules(smiles_list: str, n_clusters: int = 5) -> str:
+    """Cluster molecules based on structural similarity using Morgan fingerprints"""
+    from rdkit.Chem import AllChem
+    from sklearn.cluster import KMeans
+    import pandas as pd
+    
+    lines = [line.strip() for line in smiles_list.strip().split('\n') if line.strip()]
+    if not lines:
+        return "Please enter at least one SMILES string (one per line)"
+    
+    if len(lines) < 2:
+        return "Please enter at least 2 SMILES strings for clustering"
+    
+    # Generate fingerprints
+    mols = []
+    valid_smiles = []
+    fps = []
+    
+    for smiles in lines[:100]:  # Limit to 100 molecules
+        mol = Chem.MolFromSmiles(smiles)
+        if mol is not None:
+            fp = AllChem.GetMorganFingerprintAsBitVect(mol, 2, nBits=1024)
+            mols.append(mol)
+            valid_smiles.append(smiles)
+            fps.append(fp)
+    
+    if len(fps) < 2:
+        return "Need at least 2 valid SMILES for clustering"
+    
+    # Convert fingerprints to numpy array
+    fp_array = np.array([list(fp) for fp in fps])
+    
+    # Perform clustering
+    n_clusters = min(n_clusters, len(fps))
+    kmeans = KMeans(n_clusters=n_clusters, random_state=42, n_init=10)
+    clusters = kmeans.fit_predict(fp_array)
+    
+    # Create results
+    results = []
+    results.append(f"Clustered {len(valid_smiles)} molecules into {n_clusters} groups\n")
+    results.append("Cluster\tSMILES\tMW\tLogP")
+    
+    for i, (smiles, cluster_id) in enumerate(zip(valid_smiles, clusters)):
+        mol = mols[i]
+        mw = Descriptors.MolWt(mol)
+        logp = Descriptors.MolLogP(mol)
+        results.append(f"{cluster_id + 1}\t{smiles}\t{mw:.2f}\t{logp:.2f}")
+    
+    return "\n".join(results)
+
+
+def analyze_scaffolds(smiles_list: str) -> str:
+    """Extract and analyze molecular scaffolds (Bemis-Murcko scaffolds)"""
+    from rdkit.Chem.Scaffolds import MurckoScaffold
+    from collections import Counter
+    
+    lines = [line.strip() for line in smiles_list.strip().split('\n') if line.strip()]
+    if not lines:
+        return "Please enter at least one SMILES string (one per line)"
+    
+    scaffolds = []
+    mol_to_scaffold = []
+    
+    for smiles in lines[:100]:  # Limit to 100 molecules
+        try:
+            mol = Chem.MolFromSmiles(smiles)
+            if mol is not None:
+                scaffold = MurckoScaffold.GetScaffoldForMol(mol)
+                scaffold_smiles = Chem.MolToSmiles(scaffold)
+                scaffolds.append(scaffold_smiles)
+                mol_to_scaffold.append((smiles, scaffold_smiles))
+        except:
+            continue
+    
+    if not scaffolds:
+        return "No valid scaffolds could be extracted"
+    
+    # Count scaffold frequencies
+    scaffold_counts = Counter(scaffolds)
+    
+    results = []
+    results.append(f"Analyzed {len(mol_to_scaffold)} molecules")
+    results.append(f"Found {len(scaffold_counts)} unique scaffolds\n")
+    results.append("=== Most Common Scaffolds ===")
+    
+    for scaffold, count in scaffold_counts.most_common(10):
+        results.append(f"\nScaffold: {scaffold}")
+        results.append(f"Frequency: {count} molecules ({100*count/len(scaffolds):.1f}%)")
+        
+        # Show examples
+        examples = [smiles for smiles, scaf in mol_to_scaffold if scaf == scaffold][:3]
+        results.append("Examples:")
+        for ex in examples:
+            results.append(f"  - {ex}")
+    
+    return "\n".join(results)
+
+
 smiles_interface = gr.Interface(
     fn=smiles_to_canonical,
     inputs=gr.Textbox(label="SMILES"),
@@ -826,6 +958,51 @@ chemiscope_interface = gr.Interface(
     cache_examples=False,
 )
 
+# Property calculation interface
+properties_interface = gr.Interface(
+    fn=calculate_properties_batch,
+    inputs=gr.Textbox(
+        label="SMILES List (one per line)", 
+        placeholder="CCO\nc1ccccc1\nCC(=O)O\nCCN",
+        lines=10
+    ),
+    outputs=gr.Textbox(label="Properties (Tab-separated)", lines=15),
+    title="Batch Property Calculator",
+    description="Calculate physicochemical properties for multiple molecules. Enter one SMILES per line (max 50).",
+    examples=[["CCO\nc1ccccc1\nCC(=O)O\nCN1C=NC2=C1C(=O)N(C(=O)N2C)C"]],
+)
+
+# Clustering interface
+clustering_interface = gr.Interface(
+    fn=cluster_molecules,
+    inputs=[
+        gr.Textbox(
+            label="SMILES List (one per line)", 
+            placeholder="CCO\nc1ccccc1\nCC(=O)O\nCCN",
+            lines=10
+        ),
+        gr.Slider(minimum=2, maximum=10, value=5, step=1, label="Number of Clusters")
+    ],
+    outputs=gr.Textbox(label="Clustering Results (Tab-separated)", lines=15),
+    title="Molecular Clustering",
+    description="Cluster molecules based on structural similarity using Morgan fingerprints and K-means (max 100 molecules).",
+    examples=[["CCO\nCCCO\nCCCCO\nc1ccccc1\nc1ccc(O)cc1\nc1ccc(N)cc1\nCC(=O)O\nCCC(=O)O\nCCCC(=O)O", 3]],
+)
+
+# Scaffold analysis interface  
+scaffold_interface = gr.Interface(
+    fn=analyze_scaffolds,
+    inputs=gr.Textbox(
+        label="SMILES List (one per line)", 
+        placeholder="c1ccc(CCN)cc1\nc1ccc(CCO)cc1\nc1ccc(CCC)cc1",
+        lines=10
+    ),
+    outputs=gr.Textbox(label="Scaffold Analysis", lines=15),
+    title="Scaffold Analysis",
+    description="Extract and analyze Bemis-Murcko scaffolds from molecules (max 100).",
+    examples=[["c1ccc(CCN)cc1\nc1ccc(CCO)cc1\nc1ccc(CCC)cc1\nCCOc1ccc(CCN)cc1\nCCc1ccc(O)cc1"]],
+)
+
 
 demo = gr.TabbedInterface(
     [
@@ -833,6 +1010,9 @@ demo = gr.TabbedInterface(
         molecule_3d_interface,
         chemiscope_interface,
         orbital_interface,
+        properties_interface,
+        clustering_interface,
+        scaffold_interface,
         smiles_interface,
         smiles_to_name_interface,
         mw_interface,
@@ -844,6 +1024,9 @@ demo = gr.TabbedInterface(
         "3D Molecule Viewer",
         "Chemiscope Explorer",
         "Molecular Orbitals",
+        "Property Calculator",
+        "Molecular Clustering",
+        "Scaffold Analysis",
         "SMILES to Canonical",
         "SMILES to Name",
         "Molecular Weight",

force_rebuild.txt

@@ -1,2 +1,2 @@
 # Force rebuild
-2025-11-08 v15
+2025-11-08 v16

requirements.txt

@@ -12,3 +12,5 @@ ase
 plotly
 matplotlib
 pillow
+scikit-learn
+pandas