Addressing the problem of lysine glycation prediction in proteins via Recurrent Neural Networks

Abstract

A distinguishing feature of the metabolic disorder diabetes involves elevated damage to cellular proteins. The primary form of alteration arises from the chemical interaction between glycating agents such as methylglyoxal and proteinaceous arginine/lysine residues, causing structural and functional disruptions in target proteins. In this study, a curated version of the CPLM database to implement a recurrent neural network strategy for the classification of lysine glycation has been utilized. By using one physical property for the characterization of amino acids next to lysine sites (i.e., isoelectric point), it was possible to obtain a 59.6% accuracy for correctly predicting lysine glycation. When two properties were combined, i.e., mass and torsion angle, the accuracy increased to 59.9%. Overall, this approach can aid the task of narrowing down possible sites of lysine glycation in protein targets for further analysis.