A framework for discovering important patterns through parallel mining of protein–Protein interaction network

Abstract

Association rule mining can be applied in the field of bioinformatics for identification of co-occurrences between various biological elements such as genes and protein. In bioinformatics, protein–protein interaction network provides useful information regarding the functions of protein. Association analysis has been used for identification offrequently occurring interactionsamong the proteins in the network for predicting the functions of proteins. As the amount of data is increasing exponentially, parallel implementation of association analysis for identification of co-occurrences between proteins in protein–protein interaction network will be more efficient, fast, and scalable. In this paper we proposed an efficient framework for association analysis of frequently occurring pattern in the protein–protein interaction network. The algorithm has been parallelized using Hadoop software. The performance view of the parallel algorithm has been depicted in graph and it shows that the parallel version is more effective than the sequential one.