Molecular featurization for ML (100+ featurizers). ECFP, MACCS, descriptors, pretrained models (ChemBERTa), convert SMILES to features, for QSAR and molecular ML.
Molfeat is a robust Python library designed for molecular featurization, addressing the need for effective representation of chemical structures in machine learning tasks. By converting SMILES strings or RDKit molecules into numerical formats, Molfeat supports a wide range of applications including quantitative structure-activity relationship (QSAR) modeling, virtual screening, and molecular similarity searching. Its comprehensive suite of over 100 pre-trained embeddings and custom featurizers simplifies the process of preparing molecular data for machine learning workflows.
You can install Molfeat using pip with the command 'pip install molfeat'. For all optional dependencies, use 'pip install "molfeat[all]"'.
Molfeat can handle both SMILES strings and RDKit 'Chem.Mol' objects for molecular representation.
Yes, Molfeat provides scikit-learn compatible transformers to facilitate batch processing and integration into scikit-learn pipelines.
Molfeat is primarily used for molecular machine learning, virtual screening, similarity searching, chemical space analysis, deep learning, and cheminformatics tasks.
Yes, optional dependencies are available for specific featurizers, such as 'molfeat[dgl]' for GNN models and 'molfeat[transformer]' for ChemBERTa and other pretrained models.
Quick Setup:
.claude/skills/Repository
K-Dense-AI/claude-scientific-skills