A Sequential Gene Expression Data Bi-clustering Method: Clustering and Verification

With the development of the technique of DNA chips, more and more experiments data can be gained. Genes exhibiting similar patterns are often functionally related, which is very helpful for the research on analyzing the underlying mechanisms of metabolic and regulatory networks in the cell. But the fact that the correlated part is entangled with the unrelated part in the data, and the data is usually corrupted by some biological or experimental factors adds great difficulty to detect all of the possible biclusters. The novelty of the method proposed in this paper lies in the given data is not treated as a unique one but is segmented into many 2-combinations. And clustering is done with all these 2-combinations. Then based on the clustering results some binary tables are created. The second part is to verify the concatenated quasi-biclusters. The whole processing aims at finding all of the possible biclusters from large to small. At the end of the paper, an experiment result with a simulated small but complicated data is presented.