Using Logistic Regression Method to Predict Protein Function from Protein-Protein Interaction Data

Abstract

Protein function determination is one of the most important issues in biology research. In this paper, a new method, which is based on logistic regression method, is introduced to predict protein function from protein-protein interaction data. In the proposed method, associations among different functions are taken into account by representing a protein using all the functional annotations of its interaction protein partners. We apply our method to a constructed data set for yeast based upon protein function classifications of FunCat scheme and upon the interaction networks collected from BioGrid. The results obtained by 3-fold cross-validation test show that the proposed method can obtain desirable results for protein function prediction and outperforms some existing approaches based on protein-protein interaction data.