Microarray data mining: A novel optimization-based approach to uncover biologically coherent structures

We apply our proposed iterative algorithm to three sets of experimental DNA microarray data from experiments with the yeast Saccharomyces cerevisiae and show that the proposed iterative approach improves biological coherence. Comparison with other clustering techniques suggests that our iterative algorithm provides superior performance with regard to biological coherence. An important consequence of our approach is that an increasing proportion of genes find membership in clusters of high biological coherence and that the average cluster specificity improves.The results from these clustering experiments provide a robust basis for extracting motifs and trans-acting factors that determine particular patterns of expression. In addition, the biological coherence of the clusters is iteratively assessed independently of the clustering. Thus, this method will not be severely impacted by functional annotations that are missing, inaccurate, or sparse.DNA microarray technology allows investigators to monitor simultaneously the expression behavior of essentially all the genes within an entire genome and can provide information on gene functions and transcriptional networks. However, the large number of genes and the complexity of the underlying biological networks make extracting this information a formidable task. A common first step to interpret DNA microarray data is to cluster the data on the basis of similarity of expression patterns. Since genes with similar function often show a common expression pattern, clustering genes of known functions with poorly characterized genes provides a means of gaining insights into the functions of the latter [1]. Furthermore, patterns seen in genome-wide expression data can reveal potential gene regulation networks and relate cellular processes to changes in cellular conditions [2,3]. However, the extent to which clustering reveals useful information about the system under study depends on the extent to which the clustering method succes