Automated Analysis of Molecular Simulations: can we get there?

Antonia S. J. S. Mey, School of Chemistry, University of Edinburgh

Proteins regulate and manage all processes that make up life as we know it. Understanding their structure and how their dynamics relates to their function allows us to unravel intricate and complex biological process. A better understanding of these processes can then be used to regulate them, and counteract e.g. disease. Molecular simulations can shed light onto how proteins work, but rigorous statistical methods are needed to extract meaningful information from these simulations. Can this be automated and optimised such that we can obtain quantitative data that can be compared to experimental? Markov state models are a way of analysing simulation trajectory data, but these require a lot of modelling choices.

In this talk, I will explore the good, bad, and ugly when it comes to trying to optimise and automate modelling choices for Markov state models to understand protein dynamics. For this illustrated journey into the world of Markov state modelling we will look at the role of Cyclophilin A and its mutants proline isomerisation, as well as how small peptides/toy proteins such as BBA and chignolin fold.

The seminar will be followed by drinks and snacks in G block. Everyone is welcome to come along.