The Prediction of Succinylation Site in Protein by Analyzing Amino Acid Composition

Protein Succinylation is a kind of post-translational modification (PTM) where a succinyl group is attacked to a lysine residue of a protein molecule. Recent findings have demonstrated the important role of Succinylation in not only taking part in various biological processes but also associating with many diseases. There are many practical methods to identify succinylation sites but an expensive cost and time-wasting should be considered. The lack of research in structure and characteristic of protein will limit to understand and discover significantly. Therefore, this work aims to focus on develop a bioinformatics method for investigating Succinylation site based on the amino acid composition and physicochemical properties. Various features were investigated in this study, including 20 Binary coding, amino acid composition (AAC), amino acid pair composition (AAPC), solvent-accessible surface area (ASA), amino acid substitution matrix (Blosum62), and position-specific scoring matrix (PSSM). Evaluation by five-fold cross validation indicated that the selected features were effective in the identification of Succinylation sites. The model constructed from hybrid features, including BLOSUM62 and PSSM, yielded the best performance with sensitivity, specificity, accuracy and MCC measurements of 0.66, 0.68, 0.67 and 0.32, respectively.