GlycoMinestruct: a new bioinformatics tool for highly accurate mapping of the human N-linked and O-linked glycoproteomes by incorporating structural features

Abstract

Glycosylation plays an important role in cell-cell adhesion, ligand-binding and subcellular recognition. Current approaches for predicting protein glycosylation are primarily based on sequence-derived features, while little work has been done to systematically assess the importance of structural features to glycosylation prediction. Here, we propose a novel bioinformatics method called GlycoMinestruct(http://glycomine.erc.monash.edu/Lab/GlycoMine_Struct/) for improved prediction of human N- and O-linked glycosylation sites by combining sequence and structural features in an integrated computational framework with a two-step feature-selection strategy. Experiments indicated that GlycoMinestruct outperformed NGlycPred, the only predictor that incorporated both sequence and structure features, achieving AUC values of 0.941 and 0.922 for N- and O-linked glycosylation, respectively, on an independent test dataset. We applied GlycoMinestruct to screen the human structural proteome and obtained high-confidence predictions for N- and O-linked glycosylation sites. GlycoMinestruct can be used as a powerful tool to expedite the discovery of glycosylation events and substrates to facilitate hypothesis-driven experimental studies.

iProt-Sub: a comprehensive package for accurately mapping and predicting protease-specific substrates and cleavage sites

“…Numerous studies have suggested that a high-quality, wellestablished data set is crucial for training a robust and reliable prediction model of protease cleavage sites [37,[43][44][45]. In this study, we constructed a well-prepared benchmark data set for assessing the predictive performance of our method and other existing methods.…”

“…To improve the feature representation ability and identify a subset of optimal features that contribute the most to the prediction of substrate cleavage sites, we used a two-step feature selection strategy, which combined mRMR [42] with forward feature selection (FFS) as described in our previous work [30,44,45,94].…”

PRISMOID: a comprehensive 3D structure database for post-translational modifications and mutations with functional impact

“…Previous studies have shown that 3D structure-based features play an important role in improving the predictive performance of PTM sites [38,50]. Therefore, we calculated two commonly used protein 3D structure features, secondary structure features and solvent accessible area features by using DSSP [42] and NACCESS [43], respectively.…”

“…Since solvent accessible area (ASA) has been widely used in many bioinformatics fields, including hot spots identification [51], catalytic sites prediction [52], and glycosylation sites prediction [38]…”

“…On the whole, there is a generalized lack of user-friendly tools for both biologists and bioinformaticians to investigate the sequence-structural-function information of PTM proteins from existing protein sequence-based databases, limiting research efforts. For instance, although there are many bioinformatics prediction tools based on protein sequence data to investigate PTMs [4,[33][34][35][36][37], there are few prediction studies using PDB structures to build prediction models [38][39][40]. This is mainly due to the fact that the process of mapping from protein sequences to protein structures is quite time-consuming and operationally complex.…”