Adding tracks to aid analysis with GENS

[alieden]

As previously discussed at the meeting 12/9 we would like to improve the analysis of SVs using both GENS and vcf2cytosure. When it comes to GENS we would like to add a track with common CNVs (data from our clinical array-cgh database). Gains should be colored blue or possibly green and losses should be red. It should also be possible to see the frequency (provided in the attached file). Final_common_CNV_clusters_0.9_jaccard_0.002_freq_adusted_211207.txt

[dnil]

Overlaps issue #25, but good info, and perhaps better like this for prioritisation.

[raysloks]

Hey, sorry for taking so long with this.

Some minor issues:

• Annotation tracks are currently global, which I think was mentioned at the meeting.
• Had to add the default annotation track "databaseExport_220512", otherwise the dropdown menu wouldn't work. I uploaded a dummy version of it.
• Gens currently only cares about a select few .bed-columns when loading annotations, so I elected to merge Gain/Loss and Freq into Name.

I've uploaded the file with its name as is, but any name not containing a . should work iirc.

Edit: Added color based on Gain/Loss as suggested.

[dnil]

Hey, sorry for taking so long with this.

Thank you! Life is life; priorities eh!

Some minor issues:

• Annotation tracks are currently global, which I think was mentioned at the meeting.

Thats fine, may be something we need to think about when the balsamic crew wants more cases, but for now all good.

• Had to add the default annotation track "databaseExport_220512", otherwise the dropdown menu wouldn't work. I uploaded a dummy version of it.

Right, we specify it here: https://github.com/Clinical-Genomics/servers/blob/518e790ffcbded1bd2a8c0c053e99d9615bbd48c/config/CGVS/config/app/cg-prod-services.scilifelab.se/gens.conf.py#L8

Perhaps this cluster track is a better starting point for now; we could switch the default for it on the servers repo.

• Gens currently only cares about a select few .bed-columns when loading annotations, so I elected to merge Gain/Loss and Freq into Name.

Good call, and with the color rename. We can think about extending that, but one thing at the time.

I've uploaded the file with its name as is, but any name not containing a . should work iirc.

Ping to @alieden to see if he has a good alias? "Common_CNV_clusters_211207"?

Looks all good otw; please document somewhere, either in Atlas, or in a PR here, then perhaps including a small script to do the bed-file color preprocessing plus commands to load to stage and prod to local docs. Dunno, something like:

On hasta, activate the gens stage env and load your processed bed file:

conda activate S_gens
gens load annotations -f my_processed_bed.bed -b 37

If all tests look nominal on the gens stage server, load to production.

conda activate P_gens
gens load annotations -f my_processed_bed.bed -b 37

[dnil]

Some docs have been added to the readme with #49, and prod tracks loaded.