Implementation of regularized Markov clustering algorithm on protein interaction networks of schizophrenia's risk factor candidate genes

Abstract

Schizophrenia has been suffered by over 21 million people worldwide. Genetic and environmental issues are one of the contributing factors in the development of this disease. Some research shown that several related genes may increase the risk of this disease. Candidate genes that obtained from several research turns up linked in a large network of protein-protein interaction (PPI). Therefore, it is necessary to study the PPI network of the candidate genes. Regularized Markov Clustering Algorithm (RMCL) is a graph clustering method which is the modification of Markov Clustering Algorithm (MCL). RMCL process that is built using R programming language is applied to PPI networks of schizophrenias risk factors candidate genes data obtained from BioGRID database. RMCL algorithm simulation performed with different parameter of inflation. Then, the results of RMCL algorithm simulation is compared to MCL algorithm simulation with the same parameters. RMCL algorithm provides results in the form of overlapping clusters, which mean there are relation between clusters. Thus, based on the results of RMCL algorithm simulation, there are relation between protein clusters of several candidate genes, one of which is the relation of NRG1 and CACNG2 gene product.