DeepNGlypred: A deep neural network-based approach for human N-linked glycosylation site prediction

Abstract

Protein N-linked glycosylation is a post-translational modification that plays an importantrole in a myriad of biological processes. Computational prediction approaches serve as comple-mentary methods for the characterization of glycosylation sites. Most of the existing predictors forN-linked glycosylation utilize the information that the glycosylation site occurs at the N-X-[S/T] se-quon, where X is any amino acid except proline. Not all N-X-[S/T] sequons are glycosylated, thus theN-X-[S/T] sequon is a necessary but not sufficient determinant for protein glycosylation. In thatregard, computational prediction of N-linked glycosylation sites confined to N-X-[S/T] sequons is animportant problem. Here, we report DeepNGlyPred a deep learning-based approach that encodes thepositive and negative sequences in the human proteome dataset (extracted from N-GlycositeAtlas)using sequence-based features (gapped-dipeptide), predicted structural features, and evolutionaryinformation. DeepNGlyPred produces SN, SP, MCC, and ACC of 88.62%, 73.92%, 60%, and 79.41%,respectively on N-GlyDE independent test set, which is better than the compared approaches. Theseresults demonstrate that DeepNGlyPred is a robust computational technique to predict N-Linkedglycosylation sites confined to N-X-[S/T] sequon. DeepNGlyPred will be a useful resource for theglycobiology community