Version Compatibility

Reference examples tested with: RDKit 2024.09+, pharmer / pharmit (web service), PharmIT 1.1+, plip 2.4+ (interaction analysis).

Before using code patterns, verify installed versions match. If versions differ:

• Python: pip show rdkit then help(rdkit.Chem.Pharm3D) to check signatures

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

Pharmacophore Modeling

Hard rules

1. No fabricated citations. Every cited work must resolve to a verifiable
2. No claim without provenance. Every quantitative or factual claim
3. No silent failure. Every script invocation, API call, or tool use must declare its exit status and what to do on non-zero. A skill that silently swallows errors is a violation.

When to use

Load this skill when the user asks a question that matches its declared trigger conditions (see the frontmatter description for the most common ones). Do not load it for questions outside its scope — defer to the appropriate ORS skill instead.

When NOT to use

Do not load this skill if the question is in a sibling skill's domain (see ## Cross-references), if the user explicitly asks for a different tool, or if the task is outside the skill's declared category.

Pharmacophore Feature Types

Tolerances are pharmacophore-feature distance windows in the search. Tighter tolerances = fewer hits but more specific.

Method Taxonomy

Decision Tree by Scenario

Ligand-Based Pharmacophore (RDKit Pharm3D)

Goal: Extract a common pharmacophore from a set of bioactive compounds.

Approach: Embed actives, extract per-molecule features, then build a target pharmacophore with distance bounds capturing cross-active variability.

from rdkit import Chem
from rdkit.Chem import AllChem, ChemicalFeatures
from rdkit.Chem.Pharm3D import Pharmacophore
from rdkit.RDPaths import RDDataDir
import os

fdef_file = os.path.join(RDDataDir, 'BaseFeatures.fdef')
factory = ChemicalFeatures.BuildFeatureFactory(fdef_file)

active_smiles = ['CC(C)c1ccc(C(=O)NCc2ccccn2)cc1',
                 'CCC(C)c1ccc(C(=O)NCc2ccccn2)cc1']
active_mols = [Chem.AddHs(Chem.MolFromSmiles(s)) for s in active_smiles]
for m in active_mols:
    AllChem.EmbedMolecule(m, AllChem.ETKDGv3())
    AllChem.MMFFOptimizeMolecule(m)

feature_lists = [factory.GetFeaturesForMol(m) for m in active_mols]

# Example 2-feature pharmacophore (Aromatic + Donor) with a 3.5-5.0 A band
feature_types = ['Aromatic', 'Donor']
pharmacophore = Pharmacophore.Pharmacophore(feature_types)
pharmacophore.setLowerBound(0, 1, 3.5)
pharmacophore.setUpperBound(0, 1, 5.0)

BaseFeatures.fdef (RDKit-shipped) defines feature SMARTS. For drug-like pharmacophores, this is the standard starting point.

Receptor-Based Pharmacophore (apo2ph4 workflow)

Goal: Derive a pharmacophore from a protein binding-pocket structure without requiring a bound ligand.

Approach: Identify hot-spots (donor / acceptor / hydrophobe regions) from protein geometry; assemble into a pharmacophore. apo2ph4 (Heider et al 2023) describes this pipeline.

# Conceptual apo2ph4-style workflow; verify the exact CLI against the published release.
apo2ph4 -pdb receptor.pdb \
        -site_residues 'A:100,A:101,A:104,A:108' \
        -output pharmacophore.ph4

When a co-crystal ligand is available, derive pharmacophore directly from the ligand binding pose: each ligand feature in contact with a complementary protein residue is part of the pharmacophore.

from plip.basic import config

PLIP (Protein-Ligand Interaction Profiler) extracts per-residue interactions; use to build the pharmacophore manually from contacts (H-bond, hydrophobic, salt bridge, π-stacking).

Pharmer / Pharmit (Web Service)

For library screening with a .ph4 pharmacophore:

• Pharmer (Koes 2012): open-source pharmacophore search; efficient indexing
• Pharmit (Sunseri 2016): web-based frontend, integrates Pharmer with vendor catalogs (Enamine, Mcule, ZINC)

# Pharmit provides a REST API for batch search
curl -X POST https://pharmit.csb.pitt.edu/api/search \
     -H 'Content-Type: application/json' \"
     -d '{"pharmacophore": "...", "library": "enamine-real", "max_hits": 1000}'

For open-source Pharmer, install via conda; programmatic interface in Python.

Diffusion-Based Pharmacophore Generation (PharmacoForge)

PharmacoForge (Flynn et al 2025) generates molecules conditioned on a pharmacophore using a diffusion model. Use when de novo design is needed and a pharmacophore constraint is desired.

# PharmacoForge pseudo-API; verify against the published release.
# from pharmaco_forge import generate
# mols = generate(
#     pharmacophore='ph4_file',
#     n_molecules=1000,
#     scaffold_hint='aryl_sulfonamide',
# )

Per-Tool Failure Modes

Ligand-based pharmacophore -- too few actives

Trigger: Active set < 5 compounds with diverse chemotypes.

Mechanism: Common pharmacophore requires shared features; small active sets don't converge.

Symptom: All-atom pharmacophore is too permissive; thousands of hits with random features.

Fix: Use receptor-based pharmacophore if crystal available; or pick 1 bioactive conformer as single template.

apo2ph4 -- ambiguous pocket definition

Trigger: Binding site contains multiple sub-pockets; user-specified residues span all.

Mechanism: Hot-spot detection over generalizes.

Symptom: Pharmacophore has too many features; nothing matches in real library.

Fix: Manually curate sub-pocket; pass focused residue list (3-6 residues in active site).

Pharmer / Pharmit -- too few hits

Trigger: Pharmacophore strict; library small.

Mechanism: Geometric tolerance + feature count filter compounds aggressively.

Symptom: < 10 hits returned for large library.

Fix: Increase tolerance by 0.5-1.0 Å; remove 1 least-essential feature; verify pharmacophore geometry.

Pharmacophore match but binder is inactive

Trigger: Compound matches pharmacophore but IC50 > 10 µM.

Mechanism: Pharmacophore captures necessary features but not sufficient (strain, dynamics, off-target).

Symptom: False positive in screen.

Fix: Combine pharmacophore with shape (ROCS color) and physchem filters.

Common Errors

References

• Heider et al., J. Chem. Inf. Model. 63:147-158 -- apo2ph4 receptor-based pharmacophore.
• Flynn et al., Front. Bioinform. -- PharmacoForge diffusion-based generation.
• Koes et al., J. Chem. Inf. Model. 52:1159 -- Pharmer.
• Sunseri et al., J. Mol. Graph. Model. -- Pharmit web service.
• Wermuth et al. -- IUPAC pharmacophore definition.

Related Skills

• chemoinformatics/scaffold-analysis - 2D scaffold analysis
• chemoinformatics/shape-similarity - 3D shape
• chemoinformatics/similarity-searching - 2D similarity
• chemoinformatics/virtual-screening - Docking as alternative
• chemoinformatics/conformer-generation - 3D conformers for active alignment
• chemoinformatics/generative-design - Generative models with pharmacophore conditioning

Cross-references

Other skills in this category:

• admet-prediction
• conformer-generation
• covalent-design
• docking-rescoring
• free-energy-calculations
• generative-design
• molecular-descriptors
• molecular-io
• molecular-standardization
• pose-validation
• protac-degraders
• qsar-modeling
• retrosynthesis
• scaffold-analysis
• shape-similarity
• similarity-searching
• substructure-search
• virtual-screening

Changelog

• 1.1.0 (migration) — Bulk-migrated to v0.4.0 schema: canonical metadata block, base Hard rules, Cross-references. Body content unchanged; author should review and fill in any domain-specific extensions to the Hard rules.
• 1.0.0 — Initial release.