A dynamic look into hydrogen bonding In proteins

Abstract

This thesis aims to highlight the importance of analysing the dynamic nature of hydrogen bonds and their temporal variation. We also evaluate whether distance alone can and should be used as the main estimator for hydrogen bond strength. Molecular dynamics simulations of three common protein building blocks were conducted in triplicate to provide temporal data for our investigation. This data was then run through a newly developed program called Hybond, which analyses a given hydrogen bonding interaction as a function of time based on the reduced electron density gradient. The results found were promising and showed that the strength of a given hydrogen bond changes significantly throughout each simulation trajectory. This demonstrates the dynamic nature of these hydrogen bonds, and we provide a workflow to efficiently characterise individual interactions as a function of time. We found that distance correlates well with kinetic energy density however there are points that do not fit this correlation. In these cases, hydrogen bond strength cannot be estimated through distance alone. Other properties that Hybond outputs have also shown promising correlations with kinetic energy density, and these can be used in place of distance to estimate the strength of the interaction. Finally, we investigated some of the challenges that arise when trying to determine the time-averaged strength of a given hydrogen bond and how different averaging approaches might be more appropriate than others, depending on the underlying science question that is being asked.