EMS: Evolution-based Motif Search

APPLICATIONS / DATABASES

SUPPLEMENTAL DATA

COLLABORATIVE PROJECTS

Home > Applications > EMS

EMS: Evolution-based Motif Search

QUERY | FORMAT | CONTACT US | LINKS | CREDITS

Brief Introduction
EMS scans conserved cis-regulatory elements in human/mouse based on cross-species comparison. Sequences and alignments of all human and mouse orthlogous gene pairs are stored for searching. The list of human/mouse orthlogous genes was generated based on NCBI Homologene (version: May 2004). From the UCSC goldenpath database, (version: human, July, 2003 (hg16), 21231 Refseq genes; mouse, Feb. 2003 (mm4), 17724 Refseq genes), only full-length genes, which transcriptional start sites are ahead of the transnational start sites are kept. Finally, 11836 pairs of ortholog genes appear in both datasets, Then we extracted their 10k (up or down-stream) RepeatMasked upstream regions and aligned them using the standalone AVID tool. (See the Flowchart) When searching user-defined matrices (or in other formats like Fasta), which may generate from a cluster of genes with a same expression patter in Microarray analyses, EMS first scans them on all 11836 up-stream regions of human genes for significant candidates, then matches the mouse corresponding regions, and finally reports conserved hits (See the paper). EMS website at St. Jude uses the Hartwell Center's 280-CPU Linux cluster for performing motif scans in parallel. With this computational power, a full search takes only around 2 minutes. And a human chromosome viewer is also available to show the genome location of hits.

Query Submission (motif width>=6) see an example or see precomputed motifs from Transfac

Input a Matrix Or alignment (raw) Or alignment (Fasta)

Upstream 10k Or downstream 10k

Parameters: show top hits (max: 100)

Format example

Any of the three formats below can be used as input. In the matrix example, note that the sums of the scores in each row need to be equal. Line 01's sum (6 + 0 + 1 + 0) equals Line 02's sum (1 + 0 + 6 + 0).

1. Matrix (TransFac)	2. alignment	3. alignment (Fasta)
01 6 0 1 0 A 02 1 0 6 0 G 03 3 0 0 4 W 04 5 0 2 0 A 05 0 5 2 0 C 06 4 1 1 1 A 07 0 1 1 5 T 08 1 1 3 2 N 09 3 1 0 3 W 10 2 0 0 5 T 11 0 0 7 0 G 12 0 0 0 7 T 13 2 0 0 5 T 14 0 7 0 0 C 15 1 0 0 6 T	ggaacagagtgtact ggaacattatgttct gggacacggtgtgct ttcacatgatgttcc ggaccgatgagtcct ggcacatggtgtaca gggtcagggtgttcc ggaacgtgatgttct ggttcacgatgtaat gggactggacgttat ggtacggactgtgct ggatcaggacgttct gggacaggctgtgct agcacgacgcgttct ggttcatagtgatcc	>seq1 ggaacagagtgtact >seq2 ggaacattatgttct >seq3 gggacacggtgtgct >seq4 ttcacatgatgttcc >seq5 ggaccgatgagtcct >seq6 ggcacatggtgtaca >seq12 gggtcagggtgttcc >seq11 ggaacgtgatgttct >seq13 ggttcacgatgtaat >seq14 gggactggacgttat >seq17 ggtacggactgtgct >seq21 ggatcaggacgttct >seq29 gggacaggctgtgct

Contact Us
Tao.Xie@StJude.org (Dr. Tao Xie)

Reference

EMS has not been published yet, but has been presented at a conference:

Xie T, Lin B, Hood L, and Naeve C, EMS: Evolution-based Motif Search, Poster, Molecular Medicine Tri-Conference, San Francisco, California USA, Mar. 23-26, 2004.

Thanks
TFBS for drawing sequence logos!

Additional Links
ConSite Footprinter rVISTA

Last updated on: May 26, 2004