DynaSig-ML package

DynaSig-ML ("Dynamical Signatures - Machine Learning") is a Python package allowing the easy study of dynamics-function relationships in biomolecules.

Online documentation, step-by-step tutorial and installation guide: https://dynasigml.readthedocs.io
Companion repository for the tutorial: https://github.com/gregorpatof/dynasigml_mir125a_example

If you use DynaSig-ML, please cite these two articles:
https://doi.org/10.1101/2022.07.06.499058
https://doi.org/10.1093/bioinformatics/btab189

DynaSig-ML automatically computes and stores Dynamical Signatures of sequence variants using ENCoM (Elastic Network Contact Model), a sequence-sensitive coarse-grained normal mode analysis model. These Dynamical Signatures, along with optional additional data (for example the ddG of folding of the mutation), are then used to automatically train LASSO multilinear regression, in addition to any number of user-specified machine learning models, provided thay are implemented in sckikit-learn. The LASSO coefficients are automatically mapped back on the biomolecules' structure and can be easily visualized using PyMOL, leading to biological insights.

The guide provides examples using experimental miR-125a (a human microRNA) maturation efficiency data, as described in our previous work (https://doi.org/10.1371/journal.pcbi.1010777). The method is generalizable to any biomolecule on which mutational data exists, and for which an input structure is known or can be predicted with confidence.