Skip to main navigationSkip to main content
The University of Southampton
Biological Sciences

Research project: Integrated in silico prediction of protein-protein interaction motifs

Currently Active: 

Many vital protein-protein interactions are mediated by Short Linear Motifs (SLiMs) which are short proteins typically 5-15 amino acids long containing only a few positions crucial to function. This BBSRC-funded project integrates a number of leading computational techniques to predict novel SLiMs and add crucial detail to protein-protein interaction networks.

Short Linear Motifs (SLiMs) are functional microdomains of 5-15 amino acids that mediate numerous critical protein-protein interactions (PPI). The most successful de novo SLiM discovery methods are built on an explicit model of convergent evolution, identifying over-represented motifs in unrelated proteins that share a common interaction partner. Of these methods, SLiMFinder, which estimates the statistical significance of returned motifs, has the best performance on benchmark data. SLiM-mediated interactions fall into a class of PPI sometimes referred to as Domain-Motif Interactions (DMI), in which a globular, structured ('Domain') region binds a short linear ('Motif') region.

Recently, methods have been developed to identify new DMI (as opposed to domain-domain interactions) directly from high resolution 3D structural data using structural characteristics of DMI to identify short linear peptide regions that bind a globular domain. We have developed an extension to SLiMFinder, 'Query' SLiMFinder (QSLiMFinder), which can utilise this additional information explicitly to look for a motif shared between this short 'query' peptide sequence and a set of proteins that interact with the same (domain-containing) protein as the query. This substantially reduces the search space and increases the sensitivity.

By combining these methods, this project aims to identify sites on proteins that are critical for their interactions; we will mine all available 3D and PPI data (from PDB and various PPI databases) to generate a resource of predicted SLiMs, complete with proposed sites of action. In collaboration with the Eukaryotic Linear Motif (ELM) database, known True Positives will be used to benchmark analyses and develop a theoretical framework in which predictions can be assessed. New discoveries and/or improved annotation arising from this benchmarking - including new occurrences of known ELMs - will in turn be added back into the ELM resource.

This project is funded by BBSRC New Investigator Award BB/I006230/1. (Find out more on the BBSRC website) June 2011 - June 2014.



Related research groups

Molecular and Cellular Biosciences

Key Publications

Share this research project Share this on Facebook Share this on Twitter Share this on Weibo
Privacy Settings