Discovering Coherent Biclusters from Microarray Gene Expression Data

A. Mukhopadhyay (Germany), U. Maulik, and S. Bandyopadhyay (India)


Biclustering, microarray gene expression data, scaling bicluster, mean squared residue.


Biclustering algorithms aim to identify a subset of genes that are co-regulated in a subset of experimental conditions in microarray gene expression data. Mean squared residue is a popular measure that is optimized by several biclus tering algorithms to discover biclusters from the gene expression data. In a recent work, it has been shown that mean squared residue is only good in capturing constant and shifting biclusters. However scaling and coherent biclusters cannot be detected using this metric. In this article a transformation method has been proposed that transforms any type of biclusters into constant or shifting biclusters which can thereafter be easily detected using mean squared residue based biclustering algorithms. The proposed technique is utilized to modify some recent biclustering algorithms that minimize the mean squared residue. The effectiveness of the proposed technique has been proved theoretically and validated by experimentation.

Important Links:

Go Back