Skip to content

MultiMolecule

Accelerate Molecular Biology Research with Machine Learning

Codacy - 代码质量 Codacy - Coverage CodeCov - Coverage

PyPI - Version PyPI - Python Version Downloads Statistics

License: AGPL v3

Introduction

Welcome to MultiMolecule (浦原), a foundational library designed to accelerate scientific research in molecular biology through machine learning. MultiMolecule provides a comprehensive yet flexible set of tools for researchers aiming to leverage AI with ease, focusing on biomolecular data (RNA, DNA, and protein).

Overview

MultiMolecule is built with flexibility and ease of use in mind. Its modular design allows you to utilize only the components you need, integrating seamlessly into your existing workflows without adding unnecessary complexity.

  • data: Smart Dataset that automatically infer tasks—including their level (sequence, token, contact) and type (classification, regression). Provides multi-task datasets and samplers to facilitate multitask learning without additional configuration.
  • datasets: A collection of widely-used biomolecular datasets.
  • module: Modular neural network building blocks, including embeddings, heads, and criterions for constructing custom models.
  • models: Implementation of state-of-the-art pre-trained models in molecular biology.
  • tokenisers: Tokenizers to convert DNA, RNA, protein and other sequences to one-hot encodings.

Installation

Install the most recent stable version on PyPI:

Bash
pip install multimolecule

Install the latest version from the source:

Bash
pip install git+https://github.com/DLS5-Omics/MultiMolecule

Citation

If you use MultiMolecule in your research, please cite us as follows:

BibTeX
@software{chen_2024_12638419,
  author    = {Chen, Zhiyuan and Zhu, Sophia Y.},
  title     = {MultiMolecule},
  doi       = {10.5281/zenodo.12638419},
  publisher = {Zenodo},
  url       = {https://doi.org/10.5281/zenodo.12638419},
  year      = 2024,
  month     = may,
  day       = 4
}

License

We believe openness is the Foundation of Research.

MultiMolecule is licensed under the GNU Affero General Public License.

Please join us in building an open research community.

SPDX-License-Identifier: AGPL-3.0-or-later