Sequence representations and their utility for predicting protein-protein interactions

Abstract

Protein-Protein Interactions (PPIs) are a crucial mechanism underpinning the function of the cell. Predicting the likely relationship between a pair of proteins is thus an important problem in bioinformatics, and a wide range of machine-learning based methods have been proposed for this task. Their success is heavily dependent on the construction of the feature vectors, with most using a set of physico-chemical properties derived from the sequence. Few work directly with the sequence itself. Recent works on embedding sequences in a low dimensional vector space has shown the utility of this approach for tasks such as protein classification and sequence search. In this paper, we extend these ideas to the PPI task, making inferences from the pair instead of for the individual sequences. We evaluate the method on human and yeast PPI datasets, benchmarking against the established methods. These results demonstrate that we can obtain sequence encodings for the PPI task which achieve similar levels of performance to existing methods without reliance on complex physico-chemical feature sets.