CBOS-Clustering Base on the Score for Motif Discovery in Biological Network

In recent years, extensive research found that although a series of relationships between vertices in large-scale biological networks seemingly erratic, there are many frequently occurring sub-structures, and the number of these sub-structures is significantly more than that appears in randomly generated network. Experiments show that these sub-structures often have very important biological significance, such structure is defined as a motif. Traditional motifs discovery methods are used to identify exact motifs and biological networks often have inherent uncertainty and dynamic properties. Therefore, discovering the consensus motifs which have stochastic properties will become an important research direction of biological networks research. In this paper, we present a new method named CBOS (Clustering Base On the Score) which clustering sub graphs base on the score defined by combine the cluster's weight with the mismatch value between clusters to discovery consensus motifs in biological networks, because the cluster with high weight is considered to have more possibility to be defined as a motif. We apply this approach to the transcriptional regulatory networks of Escherichia coli and Saccharomyces cerevisiae, the results show that the method of CBOS can mining the consensus motifs efficiently, and compared with the existing algorithms, CBOS algorithm has a better performance.