Another question: I see your docs on themes, which works well. However, are more public examples of specific color combinations people have figured out work well? For example, is there a stock jbrowse blue theme, red theme, etc? Or perhaps examples of what color sets people have used?
Category for JBrowse abstract. submit in which category? "Application of bioinformatics to cancer biology” or "New software tools for data analysis”? I feel like it should be the former, since it doesn’t really do “analysis”, it just helps users look at stuff to do their own analysis. Anybody want to disagree?
Also, sorry working on this so late: it’s been a busy day and I had to take the dog to the vet which has really messed up my schedule.
added to cal:)
$ jbrowse add-assembly test_data/volvox/volvox.fa --out t1 --load copy
Added assembly "volvox" to t1/config.json
$ jbrowse add-track test_data/volvox/volvox-sorted.bam --indexFile=test_data/volvox/volvox-sorted.bam.csi --out t1 --load copy
Added track "volvox-sorted" to t1/config.json
$ ls -1 t1
config.json
volvox.fa
volvox.fa.fai
volvox-sorted.bam
volvox-sorted.bam.csi
using jbrowse cli 1.5.1
Maybe this line: https://github.com/GMOD/jbrowse-components/blob/24b45e29b681277e68ec088034f3ef25bb80c389/products/jbrowse-cli/src/commands/add-track.ts#L391 should be
index: index || ${fileName}.bai instead, like the part for bam?
possible issue about estimating bytes downloaded, could be possibly a method for stats estimation GMOD/jbrowse-components#2251
Pangenome hackathon+talks Dec 9-10 https://twitter.com/erikgarrison/status/1465346595835326497?t=NnmbhvRONJYJoxYX16s9tg&s=19
it's free so why not
Hello - I'm hoping to get some advice. We're trying to make a plugin that filters variants in a VCF track based on whether specific samples are non-wild type. The logic is not hard (make sure the genotype of samples !== '"./." or "0/0"); however, has ~2500 samples. At least how we're doing this makes a very noticeable lag in loading. I experimented with pre-computing an array of the non-WT samples in the getFeatures() method of a custom VcfTabixAdapter; however, this code is called repeatedly whenever any rendering happens. Is there a place I could hook into the processing of VCF features that would let me efficiently pre-compute some information, and let that information persist? Alternately, is there another way I should think about approaching this problem?
that cache would be in the custom adapter
i can write an implementation sketch of it quickly, just a seec
38 replies
@cmdcolin on #2554 you said something interesting I hope you could expand on here. you mention lazy parsing of genotypes. As noted above, we have a VCF with ~2400 samples. We previously never had code that did anything with samples until the details panel was opened. We are trying to implement code that lets the user filter sites based on specific samples having non-WT genotypes, and so we're starting to access genotype data much earlier and do this for all variants. What governs whether genotypes are parsed or not?
with #2554 , if you were accessing the genotypes for filtering, it would parse the genotypes but would not necessarily serialize to json
(binary vcf format)
@bbimber if you have the adapter config, you can get the raw header string with our "CoreGetInfo" RPC call, example here: https://github.com/GMOD/jbrowse-components/blob/1d22d290b8974ad21f7ddf8787db10f631d05d8f/packages/core/ui/AboutDialog.tsx#L48-L51. A feature will also have the VCF header data stored in it that you can access, example for INFO here: https://github.com/GMOD/jbrowse-components/blob/00c4e05cca8d054ab3456423b43e1bbd0e24be8c/plugins/variants/src/StructuralVariantChordRenderer/ReactComponent.js#L32-L36