Identification of S-glutathionylation sites in species-specific proteins by incorporating five sequence-derived features into the general pseudo-amino acid composition

• Published in: Journal of theoretical biology Vol. 398; pp. 96 - 102
• Main Authors: Zhao, Xiaowei, Ning, Qiao, Ai, Meiyue, Chai, Haiting, Yang, Guifu
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 07.06.2016
• Subjects: Amino Acid Sequence; Amino Acids - chemistry; Animals; Area Under Curve; Databases, Protein; Glutathione - metabolism; Humans; Internet; Mice; Multiple Features; Reproducibility of Results; S-glutathionylated sites; Species Specificity; Species-specific analysis; Support Vector Machine; Support vector machines; Web server; Support vector machines; S-glutathionylated sites; Species-specific analysis; Web server; Multiple Features
• ISSN: 0022-5193; 1095-8541
• DOI: 10.1016/j.jtbi.2016.03.030

As a selective and reversible protein post-translational modification, S-glutathionylation generates mixed disulfides between glutathione (GSH) and cysteine residues, and plays an important role in regulating protein activity, stability, and redox regulation. To fully understand S-glutathionylation mechanisms, identification of substrates and specific S-Glutathionylated sites is crucial. Experimental identification of S-glutathionylated sites is labor-intensive and time consuming, so establishing an effective computational method is much desirable due to their convenient and fast speed. Therefore, in this study, a new bioinformatics tool named SSGlu (Species-Specific identification of Protein S-glutathionylation Sites) was developed to identify species-specific protein S-glutathionylated sites, utilizing support vector machines that combine multiple sequence-derived features with a two-step feature selection. By 5-fold cross validation, the performance of SSGlu was measured with an AUC of 0.8105 and 0.8041 for Homo sapiens and Mus musculus, respectively. Additionally, SSGlu was compared with the existing methods, and the higher MCC and AUC of SSGlu demonstrated that SSGlu was very promising to predict S-glutathionylated sites. Furthermore, a site-specific analysis showed that S-glutathionylation intimately correlated with the features derived from its surrounding sites. The conclusions derived from this study might help to understand more of the S-glutathionylation mechanism and guide the related experimental validation. For public access, SSGlu is freely accessible at http://59.73.198.144:8080/SSGlu/. •We use support vector machines classifier combined with a two-step feature selection.•Identify S-glutathionylated sites in species-specific protein.•Incorporate five sequence-derived features into the general pseudo amino acid composition.