SNOSite: exploiting maximal dependence decomposition to identify cysteine S-nitrosylation with substrate site...

Area of Science:

• Biochemistry
• Proteomics
• Bioinformatics

Background:

• S-nitrosylation is a vital protein post-translational modification regulating protein function, localization, and stability.
• The substrate specificity governing cysteine S-nitrosylation remains largely uncharacterized, hindering a deeper understanding of its regulatory roles in cell signaling.

Purpose of the Study:

• To investigate the sequence and structural determinants of S-nitrosylation sites.
• To develop a computational model for predicting S-nitrosylation sites.
• To create a user-friendly tool for identifying S-nitrosylation sites in uncharacterized proteins.

Main Methods:

• Analysis of 586 experimentally identified S-nitrosylation sites from mouse endothelial cells.
• Utilized Maximal Dependence Decomposition (MDD) for conserved motif discovery.
• Employed Support Vector Machine (SVM) for predictive model generation and validated using five-fold cross-validation.

Main Results:

• Identified significant conserved motifs associated with S-nitrosylation sites based on amino acid composition, accessible surface area, and physicochemical properties.
• Achieved a high accuracy of 0.902 with the MDD-clustered SVM model in five-fold cross-validation.
• Successfully validated the model by accurately identifying known S-nitrosylation sites in DDAH1 and HBB proteins.

Conclusions:

• The developed computational approach effectively predicts S-nitrosylation sites by integrating sequence and structural features.
• The findings provide insights into the substrate specificity of S-nitrosylation.
• A web-based tool, SNOSite, is now available for identifying S-nitrosylation sites in novel protein sequences.