Histone Modification Database For Positive And Negative

Question

[Vikas Bansal]

Hi,

I want to know the sites for Histone modification (let's say H3K36me3) and also the site which has been proved to be not associated with this modification. Is there any source for this? (Human or Mouse)

Thanks and best regards,

Vikas

Answer 1

[Ryan Dale]

You could try ENCODE data from UCSC. You can access this via the Tables page on the UCSC Genome Browser site. For human, on the Table page try choosing the "Regulation" group, the "Broad Histone" track and then pick your favorite cell type and histone modification in the table dropdown.

You can download the full signal or called peaks/domains. To work out your analysis you might want to download a few different cell lines' worth of data in BED format, then use the BedTools multiinter program to do the multi-way intersection. For the full genome-wide data for many cell types, you'll want to go to the Downloads page to get the data in bulk.

After running multiinter on enough cell lines' data, pulling out the regions covered by the histone mod in any cell line, and running BedTools' complement program, you may be able to convince yourself that a histone mod is "not associated" with particular region -- or at least not detected using the methods employed by the ENCODE consortium, in the cell types they looked at, and with the peak-calling thresholds they used.

[Vikas Bansal]

Thanks. So there is no proved site that it is not associated with this modification. Do you know any paper which says something like - we selected this site because it is not associated with this Histone modification. So that I can have reference.

[Ryan Dale]

I'm not aware of any papers that say this specifically, but a Google Scholar search for "no [your favorite modification here]" has some potentially useful leads.

[Vikas Bansal]

Thanks a lot for your help.