A SVM based approach to identification of Gram-negative bacterial secretion system proteins

Abstract:

A SVM based approach is proposed to rapidly identify Gram-negative bacterial secretion system proteins. With the optimization feature set consisted of amino acid composition (AAC) and position specific scoring matrix (PSSM), this method adequately takes sequence and evolution information of proteins into account. Experiments show that this method has a good performance on prediction of Gram-negative bacterial secretion system proteins, which served as a useful complement to the study of bacterial secretion system.