add locuszoom data to UCSC Genome Browser?

[Maximilian Haeussler]

Hi, 

I work for the UCSC Browser and have three questions:

1) it would be possible, relatively easy, for you to add all your data as a "public track hub" to our genome browser. You could convert your files to bigBed format with "lollipop" (gwas) decorations or we could add a display mode for your existing VCF files. Our users would find it on track search and by connecting the hub on your public hubs page or via a direct link from your site. 

2) I wonder where your LD matrix file is. Sorry for the naive question, but I guess it's not visualized but mostly used for the analysis ? Is this the LD data from 1000 genomes? 

3) I wonder if we should have a track with the hg38 recombination rate data that you created a while ago, https://github.com/statgen/locuszoom-standalone/issues/1

thanks!

Max

[Andy Boughton]

Thanks for the questions!  UCSC’s browser is quite a powerful resource.

A few notes below:

1. In principle, I would like to see more integrations with other tools. For my.locuszoom.org, a challenge is that most of our datasets are private- the choice of whether to share them elsewhere is up to the person who uploaded each file. (If users do want to connect, the summary stats can be downloaded and converted manually. Summary stats for a public file- like a GLGC GWAS- can be downloaded to convert manually: https://my.locuszoom.org/gwas/533703/ -> https://my.locuszoom.org/gwas/533703/data/ )
1. What we CAN do is control the other direction, by letting people draw FROM track hubs TO our tool: we’ve recently been adding more “annotation” features (like BED tracks). 
2. As my counterpart at UCSC, could you suggest a listing of high-quality, stable track hubs that people might want to show alongside their plots?
2. Correct: our LD is drawn from 1000G. (and a slightly improved version of that data for build 38). We don’t provide separate visualization because the panel is so widely used.
1. I’d love to find a more modern public panel, as 1000G doesn’t do a great job annotating datasets with a lot of rare variants (highly imputed, etc). If you hear of one that can be used publicly without restrictions, do share!
3. As I recall, Ryan's recombination track was a direct liftover of build 37 data. Not every LZ plot includes the recombination rate line, but we like including it for our users. My general advice: if you do show it, try to provide the same features consistently across genome builds; we are seeing a gradual increase in build 38 adoption. 

-Andy Boughton

abo...@umich.edu

--
You received this message because you are subscribed to the Google Groups "locuszoom" group.
To unsubscribe from this group and stop receiving emails from it, send an email to locuszoom+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/locuszoom/92eecdbe-d652-4e7a-a59d-fbb783361d37n%40googlegroups.com.