We describe cisRED, a database for conserved regulatory elements that are identified and ranked by a genome-scale computational system (). The database and high-throughput predictive pipeline are designed to address diverse target genomes in the context of rapidly evolving data resources and tools. Motifs are predicted in promoter regions using multiple discovery methods applied to sequence sets that include corresponding sequence regions from vertebrates. We estimate motif significance by applying discovery and post-processing methods to randomized sequence sets that are adaptively derived from target sequence sets, retain motifs with p-values below a threshold and identify groups of similar motifs and co-occurring motif patterns. The database offers information on atomic motifs, motif groups and patterns. It is web-accessible, and can be queried directly, downloaded or installed locally.

No abstract

The identification of cis-regulatory elements and modules is an important step in understanding the regulation of genes. We have developed a pipeline capable of running multiple motif prediction methods on a whole genome scale.Using gene expression datasets to identify coexpressed genes and the Ensembl Compara database orthologues, we assemble input sequence sets comprised of the upstream regions of a target gene, its orthologues and co-expressed genes on the premise that such genes will share promoters by evolution (orthologues) or share regulatory control mechanisms (co-expressed genes). Co-expressed genes are identified by an approach that combines Pearson distances from multiple gene expression datasets derived from multiple experimental approaches and calibrated against the GO database. Our pipeline runs a number of established motif detection algorithms with a range of parameter settings on the input dataset. We integrate the diverse result sets by scoring motifs with a method-independent function. For each target gene, we assign p-values to the motif score by running the discovery pipeline on multiple sets of input sequence containing the target gene, non-coexpressed genes and "fake" orthologues generated by neutral numerical evolution.We have predicted 30,636 motif binding sites in human for 4,182 genes and an initial set of 472 motif binding sites in mouse for 92 genes with p < 0.001. The positive predictive value against a library of biologically confirmed regulatory sites approaches 0.4 at the highest p-value threshold.Predicted regulatory elements and other resources from the project are available at www.cisred.org.