Generating gene specific interaction profile?

Question

[peer_wuensche]

I don't know if such a question is even allowed to ask or if this is the right forum, but I was wondering if anyone could guess how the below figure was generated (the lowest -> Figure7 C). UCSC or IGV or others? Which file format would the data have to be in and which way would be the easiest to get to this format if starting from publically available data, often provided either raw or as bed/bam/bigWig etc? By now I gained some substantial R skills but lack any other language. I am more of a wet lab biologist and just started looking into the ever growing and fascinating world of bioinformatics. The image is from a Nature Genetics paper (doi:10.1038/ng.3286)

http://www.nature.com/ng/journal/v47/n6/fig_tab/ng.3286_F7.html

Appreciate any suggestions or hints!

Answer 1

[ollyburren]

Hi Peer, I can't be sure how fig 7 was generated but I suspect WashU genome browser (http://epigenomegateway.wustl.edu ) with a touch of Adobe Illustrator. This publication is relevant I believe http://www.nature.com/nmeth/journal/v10/n5/full/nmeth.2440.html. I believe Ensembl has a function to do this as well http://www.ensembl.info/blog/2015/05/27/drawing-cis-interactions-in-ensembl/. Full disclosure means that I have to reveal that I run http://www.chicp.org - but that takes a circularised rather than linear view of things that is not relevant here.

[peer_wuensche]

Thanks! I just looked into the custom track documentation of WashU, should be relatively straight forward to put in own data. And of course I love the CHICP browser! :)

Answer 2

[i.sudbery]

There is a Bioconductor library that can make this sort of plot. ... I think it's ggbio. Its been a while since I used it, but it looks like it is still maintained.

[peer_wuensche]

ggbio looks indeed like an awesome package . Not sure though if it'll do the arches on a linear chromosome view. I guess I'll stick with Olly Burrens WashU suggestion for now. In any case, thanks for pointing out ggbio!