Utilizing knowledge base of amino acids structural neighborhoods to predict protein-protein interaction sites

Abstract

BackgroundProtein-protein interactions (PPI) play a key role in an investigation of various biochemical processes, and their identification is thus of great importance. Although computational prediction of which amino acids take part in a PPI has been an active field of research for some time, the quality of in-silico methods is still far from perfect.ResultsWe have developed a novel prediction method called INSPiRE which benefits from a knowledge base built from data available in Protein Data Bank. All proteins involved in PPIs were converted into labeled graphs with nodes corresponding to amino acids and edges to pairs of neighboring amino acids. A structural neighborhood of each node was then encoded into a bit string and stored in the knowledge base. When predicting PPIs, INSPiRE labels amino acids of unknown proteins as interface or non-interface based on how often their structural neighborhood appears as interface or non-interface in the knowledge base. We evaluated INSPiRE’s behavior with respect to different types and sizes of the structural neighborhood. Furthermore, we examined the suitability of several different features for labeling the nodes. Our evaluations showed that INSPiRE clearly outperforms existing methods with respect to Matthews correlation coefficient.ConclusionIn this paper we introduce a new knowledge-based method for identification of protein-protein interaction sites called INSPiRE. Its knowledge base utilizes structural patterns of known interaction sites in the Protein Data Bank which are then used for PPI prediction. Extensive experiments on several well-established datasets show that INSPiRE significantly surpasses existing PPI approaches.

Protein–Protein Interaction Methods and Protein Phase Separation

“…About two-thirds of the tools available in the literature are completely or partially based on some machine learning algorithm leveraging the vast amount of information available in public databases, such as the PDB. However, other approaches have been described, including methods based on simple statistical inference from features (28,29,32,44), scoring functions (27), and homology transfer from templates (45)(46)(47)(48).…”

“…However, in contrast to pure sequence-based approaches, protein structural information allows researchers to fully exploit the structural neighborhood of a given residue. In this way, all features (either structure-based or sequence-based) can be aggregated by averaging over spatial nearest neighbors, and this improves prediction performance in many cases (26,27,30,32,33,48,50,51,55).…”

Development of New Methods Needs Proper Evaluation—Benchmarking Sets for Machine Learning Experiments for Class A GPCRs

“…This represents pairs of atom features together with distances between them . During the calculation of such a representation, the following steps are taken: (1) the given pair of atoms is extracted together with the shortest path between them, (2) the pairs of atoms are encoded in the form of descriptors informing about the atom types, the number of bonds in which both atoms are involved, and the topological distance between them, (3) a bit string is formed with assignment of particular position in the fingerprint (that is then set to “1”) with the use of the hashing function …”