MLCIL/scikit-fingerprints

Scikit-learn compatible library for molecular fingerprints and chemoinformatics

What it solves

It simplifies the process of converting chemical structures (SMILES strings) into numerical representations (molecular fingerprints) that can be used in machine learning pipelines. It bridges the gap between raw chemoinformatics data and production-grade ML models by providing a scikit-learn compatible interface.

How it works

The library provides a uniform .transform() API to compute various molecular fingerprints, filters, and distance measures. It leverages C++ RDKit for performance and integrates natively with scikit-learn's Pipeline and FeatureUnion, allowing users to chain molecular preprocessing directly with classifiers or regressors.

Who it’s for

It is designed for chemoinformaticians and data scientists building machine learning models for molecular property prediction, drug discovery, or chemical analysis.

Highlights

  • Extensive Fingerprint Library: Supports over 30 types of fingerprints, including ECFP, MACCS, and neural embeddings.
  • Chemoinformatics Toolset: Includes over 30 molecular filters (e.g., Lipinski Rule of 5) and 14 similarity/distance measures (e.g., Tanimoto).
  • ML Integration: Native compatibility with scikit-learn for easy hyperparameter tuning and deployment.
  • Reliability Tools: Features 11 applicability domain check methods to evaluate if a model's prediction for a new molecule is reliable.
  • Built-in Benchmarks: Provides direct access to standard datasets like MoleculeNet and Therapeutics Data Commons.

Related

  • Project
    fnnx-ai/scikit-llm

    A library that integrates Large Language Models like ChatGPT into scikit-learn, enabling enhanced text analysis tasks using a familiar ML pipeline API.

  • Project
    koaning/scikit-lego

    A collection of custom scikit-learn compatible transformers, metrics, and models to extend the functionality of machine learning pipelines.

  • Project
    feature-engine/feature_engine

    A Python library providing Scikit-learn compatible transformers for comprehensive feature engineering and selection to improve machine learning model performance.

  • Project
    sktime/skpro

    A Python library for supervised probabilistic prediction that provides tools for tabular regression, survival analysis, and uncertainty quantification with scikit-learn compatible interfaces.

  • Project
    skorch-dev/skorch

    A scikit-learn compatible neural network library that wraps PyTorch, enabling the use of scikit-learn's pipelines and grid search with PyTorch models.