Markov Clustering Algorithms and Their Application in Analysis of PPI Network of Malaria Genes

An empirical study on three graph-based clustering algorithms has been presented here. We have discussed those three clustering algorithms and made an analysis. The algorithms we have considered here are: Markov Clustering (MCL), Regularized Markov Clustering Algorithm (R-MCL), and Variable Inflation MCL (VI-MCL). We have used two types of graph networks: random network generated using synthetic data and PPI networks generated from 22 candidate genes of Malaria. We have considered the ubiquitous Dunn Index as the cluster validity index (CVI) to validate the generated clusters. Our experiments reveal that VI-MCL produces a lot of singletons set on both synthetic graph and PPI graph of those 22 genes. MCL performs best among the three algorithms we have chosen. The quality of clusters produces by R-MCL is better than that of VI-MCL.