Development of integrative methods for the structure elucidation of protein-ligand complexes combining experimental and computational data
Résumé
In the field of structural biology, the elucidation of the interactions between proteins and ligands is a crucial step in understanding their functional properties. The use of computational modelling tools[1], such as Autodock, Autodock Vina and Dock6 between others, has greatly aided in this effort, allowing for the prediction of ligand binding sites of the protein-ligand complexes. However, the accuracy of these predictions can be limited [2].
In this study, we aim to achieve a more accurate match between experimental data and computational models by using a range of data that we integrate with a program called CNS when the acquisition of precise and high quality data from X-Ray crystallography, Cryo-EM and NMR is not possible. We emulate different scenarios, from almost perfect amount and quality of information to sparse data with added noise and by using raw experimental data such as the chemical shifts with and without assignment from NMR, deuterium/proton exchange with NMR or MS, mutagenesis and more.
Our goal is bidirectional, for one side to improve the reliability and predictive power of computational modelling approaches in protein-ligand elucidation, but also limit the amount of experimental data needed to achieved a high quality complex structure.
References
[1]F. Stanzione, I. Giangreco, and J. C. Cole, “Use of molecular docking computational tools in drug discovery,” Prog. Med. Chem., vol. 60, pp. 273–343, 2021, doi: 10.1016/bs.pmch.2021.01.004.
[2]S. F. Sousa, P. A. Fernandes, and M. J. Ramos, “Protein–ligand docking: Current status and future challenges,” Proteins Struct. Funct. Bioinforma., vol. 65, no. 1, pp. 15–26, 2006, doi: 10.1002/prot.21082.