A new feature was introduced and implemented to reconstruct a feature vector in support vector machine (SVM) method to classify protein–protein interactions (PPIs) in human red blood cells (RBCs). The feature is based on the physicochemical properties of protein amino acids and requires merely the amino acid sequences of the proteins for the analysis. The accuracy of %3.87 was obtained for classification method that is comparable to similar classifications with more dimensional features. The reasonable accuracy indicated that this feature is suitable for classification of protein-protein interaction information on red blood cell proteins using support vector machine.

Keywords