Study and Analysis of Gene Expression Profile Patterns of Myocardial Infraction Dataset Using Biclustering Approach

Abstract

The proposed research work is to study and analyse the four biclustering algorithms BCPlaid BCXmotifs, BCSpectral and BCBimax using myocardial infraction microarray dataset. The methodology of the proposed research work consists of 3 Phases Pre-processing phase is used to convert the expression data profile into Binarnized form and discretized form. The need for converting the dataset into these forms helps certain algorithms to give effective result. The BCBimax algorithm takes Binarnized form of dataset as input whereas the BCXmotifs algorithm takes input as discretized form of dataset. The second phase is to study four biclustering algorithms and to analyse the experimental dataset. The biclustering results are to be validated using coherence measure value and Jaccard index. The visualization of biclustering results to be implemented using the heat map and parallel coordinates. Key Words: Data Mining, Biclustering, BCPlaid, BCXmotif, BCSpectral, BCBimax