This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How can I see the distribution of TFBS on specific enhancer in UCSC

I have the location information of TFBS and enhancer such as: chr1:851083-852146, and I would like to know the distribution of known TFBS on a specific enhancer. Whether UCSC Genome Browser can help me to identify it?

Can anyone help me?

gene tfbs ucsc
• 3,365 views • link
updated by simon.vanheeringen simon.vanheeringen • written by Jeason • Jeason

Do you want the distribution or just to know which are enriched for your enhancer regions? If it's the latter, you can try AME, which is made for just that.

• 0 views • link
written by jared.andrews07 jared.andrews07

2 answers

The UCSC Genome Browser does not offer a TFBS track, to my knowledge.

You could use FIMO to call TFBS on your reference genome, against a database like JASPAR. Or TRANSFAC, if you have a license.

Once you have TFBS in BED format, you could use BEDOPS bedmap to count or call TF hits from your FIMO scan that associate (overlap with) your regions of interest, such as your enhancer, or any number of other sites in BED format.

• 0 views • link
written by Alex Reynolds Alex Reynolds

Do you have an estimate on how long FIMO takes for hg38?

• 0 views • link
written by ATpoint ATpoint

Depends how many PWMs you want it to look for. If you use a full database, like all of the JASPAR CORE, it takes a pretty long time, but you can cut it down by breaking the genome up into chromosomes and using GNU Parallel to run it on them in parallel, then just concatenate the results. The resulting file will be very large though, so be prepared.

If you only have a few PWMs of interest, it's pretty quick. I think it remember it taking 2-3 hours, and that was without breaking up things by chromosome. Really made for scanning specific regions (like putative regulatory elements), rather than the entire genome, I think.

• 0 views • link
written by jared.andrews07 jared.andrews07

If you use a lower threshold (1e-5) a whole-genome scan is tractable and easy to do repeat set operations on with bedmap etc. Can be compressed with starch to make once and read/map many times without taking too much disk space.

• 0 views • link
written by Alex Reynolds Alex Reynolds

Thank you，This helps a lot!

• 0 views • link
written by Jeason • Jeason

You can check the ReMap UCSC track (based on ChIP-seq data):

http://tagc.univ-mrs.fr/remap/index.php?page=browse

And JASPAR 2018 has a UCSC track for predicted binding sites:

http://jaspar.genereg.net/genome-tracks/

• 0 views • link
written by simon.vanheeringen simon.vanheeringen

Log in to answer this question.

More posts like this