Using Composite Classifier Systems to Predict Protein Locations

Abstract

Predicting protein location is both an important and challenging topic in molecular and cellular biology. As we all know that the location of proteins sheds light upon the function of a protein whose location was uncertain. But the success of human genome project led to a protein sequence explosion. It is in a great need to develop a computational method for fast and reliably predicting the locations of proteins according to their primary sequences. In this paper, we use composite classifier system that was formed by a set of k-nearest neighbor (K-NN) classifiers, each of which is defined in a different pseudo amino composition vector. The location of a queried protein is determined by the outcome of voting among these constituent individual classifiers. It is show through the outcome that the classifier outperformed single classifier widely used in biological literature.