MDD-SOH

MDD-SOH predicts S-sulfenylation (sulfenic acid, -SOH) sites on cysteine residues and analyzes substrate motifs to support studies of redox regulation.

Key Features:

• Prediction of S-sulfenylation sites: Predicts S-sulfenylation (sulfenic acid, -SOH) sites on cysteine residues.
• Experimental dataset: Trained on a dataset comprising 1096 experimentally verified human proteins with S-sulfenylation.
• Amino acid composition analysis: Analyzes amino acid composition surrounding potential sulfenylation sites.
• Solvent-accessible surface area analysis: Analyzes solvent-accessible surface areas surrounding potential sulfenylation sites.
• Maximal Dependence Decomposition (MDD): Uses MDD to identify and characterize substrate motifs associated with S-sulfenylated cysteines.
• Support Vector Machine (SVM) classifier: Employs an SVM binary classifier integrating MDD-identified substrate motifs to distinguish sulfenylation versus non-sulfenylation sites.
• Model validation and performance: Validated with 5-fold cross-validation achieving an average accuracy of 0.87 and showing improved performance on an independent test set.

Scientific Applications:

• Identification of S-sulfenylated proteins: Enables computational prediction of S-sulfenylation sites on cysteine residues for proteome analysis.
• Substrate motif analysis: Reveals substrate motifs associated with S-sulfenylated cysteines using MDD.
• Redox regulation and structure-function studies: Supports investigation of redox regulation mechanisms and protein structure-function relationships mediated by S-sulfenylation.

Methodology:

Trained on 1096 experimentally verified human proteins, MDD-SOH analyzes amino acid composition and solvent-accessible surface areas around cysteine sites, applies maximal dependence decomposition (MDD) to derive substrate motifs, and trains a Support Vector Machine (SVM) binary classifier integrating MDD-identified motifs; validation used 5-fold cross-validation (average accuracy 0.87) and an independent test showing improved performance.