A study on meta heuristic approaches for finding cancer genes in microarray gene expression data

Abstract

Computational Biology for cancer investigation is a research area that contributes to the analysis of biological information using efficient and novel algorithms that can address the important issues in cancer diagnosis. Cancer identification is one amongst the foremost promising clinical research of microarray information. The most significant task in computational biology is identifying informative genes, combinations of genes and submatrix of genes which express similar behaviour among the genes and conditions with high prognostic efficiency from the microarray dataset For cancer microarray gene expression data analysis, the conventional clustering approach groups the genes over all the conditions or similarly, groups the conditions over all the genes, but in the cellular processes, a subset of genes active only under a small subset of conditions. Further, a single gene has a chance of linking more than one group and a gene may be involved in more than one biological process. Therefore it has a higher probability of finding marker genes that are associated with certain tissues or diseases. Hence a meta-heuristic biclustering approach has been explored as an alternative approach to standard biclustering techniques to identify coherent or similar patterns from gene expression datasets. newline