Predicting subcellular localization of proteins for Gram‐negative bacteria by support vector machines based on n‐peptide compositions

This method uses the support vector machines trained by multiple feature vectors based on n‐peptide compositions to predict subcellular localization for Gram‐negative bacteria, and achieves the highest prediction rate ever reported.