Predicting modulating sites on proteins

18:30 – Arrival and Catch-ups

19:00 - Predicting modulating sites on proteins: a new route to drug discovery? - by Joe Greener

Binding at one site on a protein can have an effect at a distant site. This phenomenon, known as allostery, is poorly understood despite being important in biological processes such as signalling. Allosteric drugs have advantages over conventional competitive inhibitors but remain a largely unexplored prospect. Computational methods to predict allosteric sites on proteins are necessary if the potential of allosteric drugs is to be realised.

I will describe two methods developed during my PhD to predict allosteric sites on proteins. AlloPred (http://www.sbg.bio.ic.ac.uk/allopred/home) uses normal mode analysis to model the vibrations of a target protein, and uses perturbations of the modes in a machine learning approach. It shows similar and complementary performance to existing methods and is available as a web server.

A new distance geometry approach that produces an ensemble of protein structures similar to a molecular dynamics ensemble has also been developed. A shift in the population of the ensemble due to binding at a target site can be used to predict allosteric character. Previous experimental results on model proteins have been reproduced with the method.

19:20 - Group Discussion (Opportunity to discuss any bioinformatic challenges you are facing with the group)

19:30 – Wrap up.

19:35 – Social drink in nearby pub