Dimorphite-DL and biotite-tools, two open source programs for the acceleration of structure-based drug design

[摘要] Computer-aided drug design has seen a proliferation of tools that allow the manipulationof small molecule and macromolecular structures in increasingly high-throughput settings.Molecular dynamics simulations, small molecule docking software, and visualization tools allowresearchers to rapidly identify drug candidates and narrow the list of compounds thatexperimentalists must consider for further testing. Any gap in automating computer-aided drugdesign thus delays potentially life-saving discoveries. Here we present two open-source programswe developed to address challenges facing both protein and ligand preparation. Dimorphite-DLis a lightweight python program that predicts protonation states of small molecules using anempirical approach to ensure accurate docking and modelling calculations. The presence orabsence of a hydrogen atom often determines whether a given ligand will bind a protein of interest.Biotite-tools is a python package that provides several popular statistical functions for analyzingmolecular dynamics simulations in an easy-to-use way. Conformational fluctuation is complex,and it can be challenging to extract insight from what is essentially a 'protein movie.” As such,simulation analysis has largely been restricted to those with backgrounds in computation, limitingthe scope of such a powerful tool. Biotite-tools aims to accelerate the efforts of those alreadyworking with molecular dynamics and make analysis more accessible to experimentalists.

[发布机构] the University of Pittsburgh