BiFree: An Efficient Biclustering Technique for Gene Expression Data Using Two Layer Free Weighted Bipartite Graph Crossing Minimization

Conventional clustering technique for gene expression data provides a global view of the data. In the biological prospective, a local view is essential for better analysis of gene expression data with simultaneous grouping of genes and conditions. Several biclustering techniques have been proposed in the literature based on different problem formulation. Therefore, it is difficult to compare these techniques with respect to their biological significance, effectiveness and accuracy. In this paper, we have proposed a biclustering technique based on two layer free weighted crossing minimization of a bipartite graph. Using this technique, we can mine different types of biclusters amid noise and it works well in practice for real and synthetic gene expression data. The experimental evaluation reveals the accuracy and effectiveness of this technique with respect to noise handling and execution time in comparison to other biclustering approaches. Keywords— Bipartite Graph, Crossing Minimization, Clustering, Biclustering, Gene Expression Data, Microarray.