This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Recapitulating a GenomeBrowser-style plot

I'm trying to visualize some cut&tag data, and would like to re-capitulate a plot that looks like this:

enter image description here

The key features I'm looking for are:

  1. The ability to vertically highlight regions [though I'm OK doing this in post, too]
  2. The presence of tracks without "boxes" around them -- including the gene track on the bottom
  3. The ability to show individual reads as "dots" or "dashes" (the "snow" effect)

The font for the track heights seem to suggest that this is IGV, but I haven't been able to customize it to do this (remove track boundaries, show reads as tiny dashes as opposed to the standard IGV read display). I'm open to some equivalent way of obtaining this display (in, say, WashU epigenome browser or UCSC genome browser).

Thanks

igv genome-browser
• 2,749 views • link
written by LChart LChart

I got a note today that this message got >1k views. I'm routinely generating these plots now, but not using Millefy because it is really specific to a pseudotime-style analysis.

Instead I take the idents for the appropriate Signac/Seurat cluster, spit them to a file, and create an ident-specific bigwig for each one. Typically I'll count the # of fragments and take only the top 50-100 within a specified region.

Load all the tracks into IGV, set Windowing Function -> maximum; Set Data Range... [0, 1]; Set Track Height 1. This gives 1px per track for presence/absence.

That generates the 'snow' effect in the main plot.

• 0 views • link
written by LChart LChart

thanks for updating this thread. is there any way to automate this, or is it still pretty manual? i'm sure many people would like this!

• 0 views • link
written by cmdcolin cmdcolin

I pretty much do it manually but it's automatable. You can do the sub-setting, ranking, and fragment-file splitting in an R script; call a shell script to generate the coverage tracks for each. You could also write out an IGV session programmatically too, with the boilerplate something like

<Track attributeKey="1" autoScale="false" clazz="org.broad.igv.track.DataSourceTrack" fontSize="16" height="1" id="$path" name="1" renderer="BAR_CHART" visible="true" windowFunction="max" color="$R,$G,$B">
<DataRange baseline="0.0" drawBaseline="true" flipAxis="false" maximum="1.0" minimum="0.0", type="LINEAR"/>
</Track>
• 0 views • link
written by LChart LChart

2 answers

Millefy is somewhat similar https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-020-6542-z

Signac also https://stuartlab.org/signac/articles/visualization

My list of "single cell tagged tools" for genomic data visualization https://cmdcolin.github.io/awesome-genome-visualization/?latest=true&tag=Single%20cell

I agree it looks like igv but i don't know how to make igv conjure up such a display

• 0 views • link
written by cmdcolin cmdcolin

Millefy looks great! It's close enough to what I want that I'm accepting it as the answer. Thanks!

• 0 views • link
written by LChart LChart

Have your tried PyGenomeTracks? It is a cool package to visualize any genetic data.

https://pygenometracks.readthedocs.io/en/latest/content/examples.html

• 0 views • link
written by bk11 • bk11

New to me. I don't see anything in the examples covering (3) -- do you know of a way to do this?

• 0 views • link
written by LChart LChart

I believe it should be doable.

• 0 views • link
written by bk11 • bk11

Log in to answer this question.

More posts like this