The identification of over-represented but imperfectly conserved motifs in genomic DNA is a problem with important biological applications, such as the discovery of regulatory elements that determine the timing, location, and level of gene transcription. Experimental techniques such as ChIP-chip and geneexpression

critical residues

Computational discovery of transcription factor binding site is one of the important tools in the genetic and genomic analysis. Rough prediction of gene regulation network and finding possible co-regulated genes are typical application of the technique. Countless number of the motif-discovery algorithms have been proposed for the past years. But there’re still no dominant algorithm and each algorithm doesn’t give enough accuracy without extensive information. Here we explore the possibility of combining multiple algorithms for the one integrated result for the improvement of performance and the convenience of researchers. Moreover, we apply new high-order information that is reorganized from the set of basis predictions to the final prediction.

In this paper we propose a novel approach for identification of generic motifs in an integrated manner by introducing the notion of submotifs. We formulate the motif finding problem as a constrained submotif pattern mining and present an algorithm called SPACE for identifying motifs that may contain spacers. When spacers are present, we show that the algorithm can identify motifs where 1) the spacers may be of varying lengths, 2) the number of motif segments may be unknown, and 3) the lengths of motif segments may be unknown. We perform rigorous experiments with the Motif Assessment Benchmarks by Tompa et al., and observe that our algorithm overall is able to outperform all popular algorithms tested so far, with significant improvements on sensitivity and specificity. 1.

Abstract not found

Abstract- Deciphering the complex interaction between transcriptional regulatory (both trans- and cis-) elements comprehensively and identifying these potential binding sites are fundamental problems in functional genomics. Therefore, determining the transcription factors that regulate a gene in different cell types and the cis-regulatory elements they are binding to will help lay the foundation for building gene regulatory networks. While many computational approaches have been developed for lower eukaryotes and prokaryotes, most of them often do not generalize to vertebrates. Here, we use gene ontological evidences to perform functional enrichment analysis among the TFs and genes, and group the functionally related genes to characterize their transcriptional association. We also analyze correlations between TFs and genes using their expression profiles. Thus, we search for putative transcriptional regulatory elements (transcription factor binding sites) along core promoter regions of the grouped genes. The performance of our search is highly satisfactory in term of binding site hit accuracy.

Abstract not found