Molfeat - Molecular Featurization Hub

Overview

Molfeat is a comprehensive Python library for molecular featurization that unifies 100+ pre-trained embeddings and hand-crafted featurizers. Convert chemical structures (SMILES strings or RDKit molecules) into numerical representations for machine learning tasks including QSAR modeling, virtual screening, similarity searching, and deep learning applications. Features fast parallel processing, scikit-learn compatible transformers, and built-in caching.

When to Use This Skill

This skill should be used when working with:

• Molecular machine learning: Building QSAR/QSPR models, property prediction
• Virtual screening: Ranking compound libraries for biological activity
• Similarity searching: Finding structurally similar molecules
• Chemical space analysis: Clustering, visualization, dimensionality reduction
• Deep learning: Training neural networks on molecular data
• Featurization pipelines: Converting SMILES to ML-ready representations
• Cheminformatics: Any task requiring molecular feature extraction

Does NOT Trigger

Installation

uv pip install molfeat            # molfeat 1.0.0 (current as of 2026-09); Python >= 3.11, pulls torch + datamol

# Optional extras (molfeat 1.x)
uv pip install "molfeat[transformer]"   # Hugging Face models: ChemBERTa, ChemGPT, MolT5, ...
uv pip install "molfeat[mordred]"       # Mordred descriptors (mordredcommunity)
uv pip install "molfeat[fcd]"           # FCD / ChemNet embeddings
uv pip install "molfeat[pyg]"           # PyTorch Geometric (used by Mol-JEPA)
uv pip install "molfeat[all]"

molfeat 1.0 breaking changes. The DGL-based pretrained GNNs (gin_supervised_*, jtvae_zinc_no_kl), Graphormer, and the protein featurizers were removed, along with the dgl and graphormer extras (there has never been a map4 extra); loading those model-store entries now fails. New foundation-model featurizers are CheMeleonTransformer (2,048-d, weights fetched from Zenodo and checksum-verified) and MolJEPATransformer (CC BY-NC 4.0; requires trust_remote_code=True and accept_noncommercial_license=True). If you must reproduce legacy GIN/Graphormer embeddings, pin molfeat<1 (0.11.x requires Python ≤ 3.10) in a separate environment. MAP4 needs the map4 package from https://github.com/reymond-group/map4 (not on PyPI).

Core Concepts

Molfeat organizes featurization into three hierarchical classes:

1. Calculators (molfeat.calc)

Callable objects that convert individual molecules into feature vectors. Accept RDKit Chem.Mol objects or SMILES strings.

Use calculators for:

• Single molecule featurization
• Custom processing loops
• Direct feature computation

Example:

from molfeat.calc import FPCalculator

calc = FPCalculator("ecfp", radius=3, fpSize=2048)
features = calc("CCO")  # Returns numpy array (2048,)

2. Transformers (molfeat.trans)

Scikit-learn compatible transformers that wrap calculators for batch processing with parallelization.

Use transformers for:

• Batch featurization of molecular datasets
• Integration with scikit-learn pipelines
• Parallel processing (automatic CPU utilization)

Example:

import numpy as np
from molfeat.trans import MoleculeTransformer
from molfeat.calc import FPCalculator

transformer = MoleculeTransformer(FPCalculator("ecfp"), n_jobs=-1, dtype=np.float32)
features = transformer(smiles_list)  # (n_mols, 2048) array; without dtype you get a list of arrays

3. Pretrained Transformers (molfeat.trans.pretrained)

Specialized transformers for deep learning models with batched inference and caching.

Use pretrained transformers for:

• State-of-the-art molecular embeddings
• Transfer learning from large chemical datasets
• Deep learning feature extraction

Example (PretrainedMolTransformer is the abstract base class — instantiate a concrete subclass):

import numpy as np
from molfeat.trans.pretrained import PretrainedHFTransformer

transformer = PretrainedHFTransformer(kind="ChemBERTa-77M-MLM", notation="smiles", dtype=np.float32)
embeddings = transformer(smiles_list)  # (n_mols, 384) mean-pooled embeddings

Quick Start Workflow

Basic Featurization

import numpy as np
from molfeat.calc import FPCalculator
from molfeat.trans import MoleculeTransformer

# Load molecular data
smiles = ["CCO", "CC(=O)O", "c1ccccc1", "CC(C)O"]

# Create calculator and transformer (dtype makes the output a single array)
calc = FPCalculator("ecfp", radius=3)
transformer = MoleculeTransformer(calc, n_jobs=-1, dtype=np.float32)

# Featurize molecules
features = transformer(smiles)
print(f"Shape: {features.shape}")  # (4, 2048)

Save and Load Configuration

# Save featurizer configuration for reproducibility
transformer.to_state_yaml_file("featurizer_config.yml")

# Reload exact configuration
loaded = MoleculeTransformer.from_state_yaml_file("featurizer_config.yml")

Handle Errors Gracefully

# ignore_errors is an argument of the CALL, not the constructor
# (a constructor kwarg is silently swallowed and the call still raises)
transformer = MoleculeTransformer(calc, n_jobs=-1, dtype=np.float32, verbose=True)

features, valid_ids = transformer(smiles_with_errors, ignore_errors=True)
# features: rows for the molecules that featurized; valid_ids: their input positions
# transformer.transform(smiles_with_errors, ignore_errors=True) instead keeps None placeholders

Choosing the Right Featurizer

For Traditional Machine Learning (RF, SVM, XGBoost)

Start with fingerprints:

# ECFP - Most popular, general-purpose
FPCalculator("ecfp", radius=3, fpSize=2048)

# MACCS - Fast, good for scaffold hopping
FPCalculator("maccs")

# MAP4 - Efficient for large-scale screening (needs the map4 package from GitHub)
FPCalculator("map4")

For interpretable models:

# RDKit 2D descriptors (200+ named properties)
from molfeat.calc import RDKitDescriptors2D
RDKitDescriptors2D()

# Mordred (1800+ comprehensive descriptors)
from molfeat.calc import MordredDescriptors
MordredDescriptors()

Combine multiple featurizers (FeatConcat takes fingerprint names or FPVecTransformer objects, not FPCalculators):

from molfeat.trans import FeatConcat

concat = FeatConcat(
    ["maccs", "ecfp"],                    # 167 + 2048 dimensions
    params={"ecfp": {"length": 2048}},    # FPVecTransformer's ecfp default length is 2000
    dtype=np.float32,
)
X = concat(smiles)                        # (n_mols, 2215); concat.length == 2215

For Deep Learning

Transformer-based embeddings:

# ChemBERTa - Pre-trained on 77M PubChem compounds
PretrainedHFTransformer(kind="ChemBERTa-77M-MLM", notation="smiles")

# ChemGPT - Autoregressive language model (SELFIES input)
PretrainedHFTransformer(kind="ChemGPT-1.2B", notation="selfies")

Foundation-model embeddings (molfeat 1.x):

from molfeat.trans.pretrained import CheMeleonTransformer, MolJEPATransformer

CheMeleonTransformer()                    # 2,048-d descriptor-foundation-model fingerprints
MolJEPATransformer(trust_remote_code=True,
                   accept_noncommercial_license=True)   # CC BY-NC 4.0 weights

The legacy DGL GIN (gin_supervised_*) and Graphormer models were removed in molfeat 1.0.

For Similarity Searching

# ECFP - General purpose, most widely used
FPCalculator("ecfp")

# MACCS - Fast, scaffold-based similarity
FPCalculator("maccs")

# MAP4 - Efficient for large databases
FPCalculator("map4")

# USR/USRCAT - 3D shape similarity
from molfeat.calc import USRDescriptors
USRDescriptors()

For Pharmacophore-Based Approaches

# FCFP - Functional group based
FPCalculator("fcfp")

# CATS - Pharmacophore pair distributions (189-d in 2D)
from molfeat.calc import CATS
CATS()                      # CATS(use_3d_distances=True) for the 3D variant

# Gobbi - Explicit 2D pharmacophore features
from molfeat.calc import Pharmacophore2D
Pharmacophore2D(factory="gobbi")

Common Workflows and Advanced Patterns

End-to-end recipes (QSAR model building, virtual screening, similarity search, scikit-learn pipeline integration, comparing featurizers), ModelStore discovery, and advanced usage (custom preprocessing, chunked batch processing, caching expensive embeddings) have moved to keep this body lean.

Full copy-ready workflow and advanced-pattern recipes: see references/workflows_and_patterns.md. Additional runnable examples (PyTorch training, grid search, 3D conformers) live in references/examples.md.

Performance Tips

1. Use parallelization: Set n_jobs=-1 to utilize all CPU cores
2. Batch processing: Process multiple molecules at once instead of loops
3. Choose appropriate featurizers: Fingerprints are faster than deep learning models
4. Cache pretrained models: Leverage built-in caching for repeated use
5. Use float32: Set dtype=np.float32 when precision allows
6. Handle errors efficiently: Use ignore_errors=True for large datasets

Common Featurizers Reference

Quick reference for frequently used featurizers:

*First run is slow; subsequent runs benefit from caching

Resources

This skill includes comprehensive reference documentation:

references/api_reference.md

Complete API documentation covering:

• molfeat.calc - All calculator classes and parameters
• molfeat.trans - Transformer classes and methods
• molfeat.store - ModelStore usage
• Common patterns and integration examples
• Performance optimization tips

When to load: Reference when implementing specific calculators, understanding transformer parameters, or integrating with scikit-learn/PyTorch.

references/available_featurizers.md

Comprehensive catalog of all 100+ featurizers organized by category:

• Transformer-based language models (ChemBERTa, ChemGPT)
• Graph neural networks (GIN, Graphormer — legacy, removed in molfeat 1.0)
• Molecular descriptors (RDKit, Mordred)
• Fingerprints (ECFP, MACCS, MAP4, and 15+ others)
• Pharmacophore descriptors (CATS, Gobbi)
• Shape descriptors (USR, ElectroShape)
• Scaffold-based descriptors

When to load: Reference when selecting the optimal featurizer for a specific task, exploring available options, or understanding featurizer characteristics.

Search tip: Use grep to find specific featurizer types:

grep -i "chembert" references/available_featurizers.md
grep -i "pharmacophore" references/available_featurizers.md

references/examples.md

Practical code examples for common scenarios:

• Installation and quick start
• Calculator and transformer examples
• Pretrained model usage
• Scikit-learn and PyTorch integration
• Virtual screening workflows
• QSAR model building
• Similarity searching
• Troubleshooting and best practices

When to load: Reference when implementing specific workflows, troubleshooting issues, or learning molfeat patterns.

Troubleshooting

Invalid Molecules

Enable error handling to skip invalid SMILES:

transformer = MoleculeTransformer(
    calc,
    ignore_errors=True,
    verbose=True
)

Memory Issues with Large Datasets

Process in chunks or use streaming approaches for datasets > 100K molecules.

Pretrained Model Dependencies

Some models require additional packages. Install specific extras:

uv pip install "molfeat[transformer]"  # For ChemBERTa/ChemGPT/MolT5
uv pip install "molfeat[pyg]"          # For Mol-JEPA (with the transformer extra)

There is no dgl extra in molfeat 1.x — the DGL GIN models were removed.

Reproducibility

Save exact configurations and document versions:

transformer.to_state_yaml_file("config.yml")
import molfeat
print(f"molfeat version: {molfeat.__version__}")

Additional Resources

• Official Documentation: https://molfeat-docs.datamol.io/
• GitHub Repository: https://github.com/datamol-io/molfeat
• PyPI Package: https://pypi.org/project/molfeat/
• Tutorial: https://portal.valencelabs.com/datamol/post/types-of-featurizers-b1e8HHrbFMkbun6

Part of the AlterLab Academic Skills suite.