Support vector machine-based method for subcellular localization of human proteins using amino acid compositions, their order, and similarity search

Here we report a systematic approach for predicting subcellular localization (cytoplasm, mitochondrial, nu-clear, and plasma membrane) of human proteins. First, support vector machine (SVM)-based modules for pre-dicting subcellular localization using traditional amino acid and dipeptide (i 1) composition achieved overall accuracy of 76.6 and 77.8%, respectively. PSI-BLAST, when carried out using a similarity-based search against a nonredundant data base of experimentally an-notated proteins, yielded 73.3 % accuracy. To gain fur-ther insight, a hybrid module (hybrid1) was developed based on amino acid composition, dipeptide composi-tion, and similarity information and attained better ac-curacy of 84.9%. In addition, SVM modules based on a different higher order dipeptide i.e. i 2, i 3, and