Abstract

Mechanistic details of how protein interactions are mediated are important for understanding and perturbing biological processes. However, such details (most often in the form of experimental 3D structures) are largely absent for the majority of interactions determined via interaction discovery methods. To address this gap we have developed several predictive tools during the past decade that can either produce fast models of interactions or predict the means by which two proteins interact via short linear motifs. We are applying these tools to allow the discovery of new interaction modules and to help to understand new datasets produced by next generation sequencing project or proteomics methods directed (for example) towards the discovery of post-translational modifications. In this talk I will summarise our methods and discuss our key recent applications of these methods.