IGV URL links and loading genome

[Kirill Tsyganov]

Hi there, 

I'm creating IGV URL links 

http://localhost:60151/load?file=http%3A%2F%2Fbioinformatics.erc.monash.edu%2Fhome%2Fkirill%2Fprojects%2Ftest%2FRNAseq-datat%2FbamFiles%2Fvivo1_sorted.bam&genome=http://bioinformatics.erc.monash.edu/home/kirill/refFiles/bacterial-genome.gbk&merge=true&name=vivo1

I made up path and hiden genome name.

Link work, and it loads up genome together with data (BAM file). merge=true I understand should merged several BAM files into single sesstion. And this works if I refere to in-build genome e.g mm10 (genome=mm10), but not when I specify http url for the genome location. 

is there a way to pull genome over http and still have merged sessions?

also would you tell me how to incode your property.txt file for genbank file?

This is what I have for in-build mm10

fasta=true

cytobandFile=cytoBand.txt

id=mm10

name=Mouse (mm10)

geneFile=refGene.txt

geneTrackName=Refseq genes

sequenceLocation=http://igv.broadinstitute.org/genomes/seq/mm10/mm10.fa

What would be key for genank file?

thanks

Kirill

[James Robinson]

Hi,

First, the "merge" problem is an igv bug.   Thanks for reporting it.   I will enter a git issue, or you can do so yourself if you are interested in tracking progress on this:  https://github.com/igvteam/igv/issues/new

Until a fix is available you could work around this problem by not including the genome parameter in the URL.    Users of the page could be instructed to select or load the correct genome separately.

There is no property.txt file associated with a genbank file, they are standalone files.    The "id" property from a genbank file is the "ACCESSION" property in the file, or if that is not specified the "LOCUS".   However these ids won't help with your current problem as they can't be used to load the file.

Jim

--

---
You received this message because you are subscribed to the Google Groups "igv-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to igv-help+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/igv-help/d4eadfdb-802f-4e95-8100-4fa4128a755a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Kirill Tsyganov]

Hi Jim, 

Thanks for your response. I'll open an issue on github. 

My work around for this was:

<a href='http://localhost:60151/load?genome=http%3A%2F%2Fbioinformatics.erc.monash.edu%2Fhome%2Fkirill%2Fprojects%2Fblah%2FGCF_000156855.2_ASM15685v2%2FGCF_000156855.2_ASM15685v2.genome'>Load genome</a>

AND property.txt is this

fasta=true

fastaDirectory=false

ordered=true

id=GCF_000156855.2_ASM15685v2

name=gonorrhoeae

geneFile=GCF_000156855.2_ASM15685v2_genomic.gff

sequenceLocation=GCF_000156855.2_ASM15685v2_genomic.fna

clicking on that link does redirect to to new page with this error message "ERROR Parameter "file" is required", but that's ok since genome still loads into IGV. Obvisuly I'd like to be no error and everything to work smooth. 

I can not include the genome since these are "custome" genomes i.e not availbale in IGV. I can provide a link to load FASTA file, but what about Annotation file? I haven't tried, but even if I can provide two URLs for FASTA and annotation files. I'd rather to minimize number of files researches need to load to minimise confusion. Web-start alone sometime is confusing ! :) 

Regarding property.txt file

Is there a list somewhere on the web of all possible keys=values pairs? Thanks  for `id` key description, but what about all the other keys? If I'm using genbank do I say `fasta=false` or have something like this `genbank=true`? and with `geneFile=` and `sequenceLocation=` both of that data comes from single files - genbank. Do I put that file name as a value for two different keys?

Thanks, 

Kirill

To unsubscribe from this group and stop receiving emails from it, send an email to igv-help+u...@googlegroups.com.

[James Robinson]

Hi Kirill,

We have a requirement for a "file" parameter although as you note this isn't really required.   I will remove this for the next release so you no longer see the error.

For genbank files the .genome property.txt file is not relevant,  as I said there is no property.txt file because a .genome file is not used.  The "genome" parameter in this case points to the genbank file.  That's all that you should need to do,  is that not working correctly?

For fasta based references you do not really need a ".genome" file.  You can just supply a link to the fasta file for the "genome" parameter, and one or more annotation tracks as files in the comma delimited list for the file parameter.  

In addition to the keys you have listed below there are  2 additional keys, an "alias" file which contains chromosome name aliases, and a "cytoband" file that contains ideogram annotations.   These are optional.

The zipped ".genome" format will be replaced in the next IGV release with a json equivalent with more options,  basically the structure that igv.js uses.   So the ".genome" format is almost deprecated, although backward compatibility will be maintained indefinitely.

Jim