Discriminating plasma membrane, internal membrane, and organelle membrane proteins by SVM

Abstract

Different membrane proteins are associated with the bio-membrane of distinct organelles which characterize their activities in and out of the cell based on their locale. Recognition of the cellular alcove is thus crucial and can be executed through machine learning algorithms. We implemented a Support Vector Machine(SVM) in order to reveal the type of membrane protein by taking the Mem Loci dataset. Before implementing SVM, we transformed the amino acid sequence into Pseudo amino acid composition and this yields an accuracy of 74%. For output, we have to enter the PseAAC of proteins and the screen reveal the category of membrane protein as either internal membrane protein or organelle membrane protein, or plasma membrane protein.