BaNDyT: Bayesian Network Modeling of Molecular Dynamics Trajectories

• DOI: 10.1021/acs.jcim.4c01981
• Publication Date: 2025-01-23T10:51:41Z
• Authors: Elizaveta Mukhaleva; Babgen Manookian; Hanyu Chen; Indira R. Sivaraj; Ning Ma; Wenyuan Wei; Konstancja Urbaniak; Grigoriy Gogoshin; Supriyo Bhattacharya; Nagarajan Vaidehi; Andrei S. Rodin; Sergio Branciamore
• Keywords: Dynamics

Bayesian network modeling (BN modeling, or BNM) is an interpretable machine learning method for constructing probabilistic graphical models from the data. In recent years, it has been extensively applied to diverse types of biomedical data sets. Concurrently, our ability perform long-time scale molecular dynamics (MD) simulations on proteins and other materials increased exponentially. However, analysis MD simulation trajectories not data-driven but rather dependent user's prior knowledge systems, thus limiting scope utility simulations. Recently, we pioneered using BNM analyzing protein complexes. The resulting BN yield novel fully insights into functional importance amino acid residues that modulate proteins' function. this report, describe BaNDyT software package implements specifically attuned We believe first include specialized advanced features a model. here software's uses, methods associated with it, comprehensive Python interface underlying generalist code. This provides powerful versatile mechanism users control workflow. As application example, have utilized methodology study how membrane proteins, G protein-coupled receptors, selectively couple proteins. can be used any as well polymeric materials.