MoCHI: neural networks to fit interpretable models and quantify energies, energetic couplings, epistasis, and allostery from deep mutational scanning data

Andre J Faure; Ben Lehner

doi:10.1186/s13059-024-03444-y

MoCHI: neural networks to fit interpretable models and quantify energies, energetic couplings, epistasis, and allostery from deep mutational scanning data

Genome Biol. 2024 Dec 2;25(1):303. doi: 10.1186/s13059-024-03444-y.

Authors

Andre J Faure^{1

2}, Ben Lehner^{3

4

5

6}

Affiliations

¹ Centre for Genomic Regulation (CRG), The Barcelona Institute of Science and Technology, Barcelona, Spain. [email protected].
² Current Address: ALLOX, PRBB Building, C/Dr. Aiguader, 88, 08003, Barcelona, Spain. [email protected].
³ Centre for Genomic Regulation (CRG), The Barcelona Institute of Science and Technology, Barcelona, Spain. [email protected].
⁴ University Pompeu Fabra (UPF), Barcelona, Spain. [email protected].
⁵ Institució Catalana de Recerca I Estudis Avançats (ICREA), Barcelona, Spain. [email protected].
⁶ Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, UK. [email protected].

Abstract

We present MoCHI, a tool to fit interpretable models using deep mutational scanning data. MoCHI infers free energy changes, as well as interaction terms (energetic couplings) for specified biophysical models, including from multimodal phenotypic data. When a user-specified model is unavailable, global nonlinearities (epistasis) can be estimated from the data. MoCHI also leverages ensemble, background-averaged epistasis to learn sparse models that can incorporate higher-order epistatic terms. MoCHI is freely available as a Python package ( https://github.com/lehner-lab/MoCHI ) relying on the PyTorch machine learning framework and allows biophysical measurements at scale, including the construction of allosteric maps of proteins.

Keywords: Allostery; Deep mutational scanning; Epistasis; Neural networks; Thermodynamic models.

MeSH terms

Allosteric Regulation
Epistasis, Genetic*
Humans
Machine Learning
Models, Genetic
Mutation*
Neural Networks, Computer*
Software*
Thermodynamics