Genomes not available error on IGV startup Windows v2.11.9

[Adam Clayton]

I know we're using an older version of IGV, but aside from that fact, several users on my team noticed some new errors popping up today suggesting connection issues with the genome tables (see attachments). While acknowledging these errors (pressing OK) doesn't appear to interfere with IGV functionality for some users, others are unable to open any bam files. Are there any known issues with servers on the IGV side today? Users have tried re-installing the same version of IGV and rebooting without success. Thanks for investigating.

-Adam

[igv-help]

Hi Adam,

Thanks for letting us know. There is an issue with our AWS buckets where some of the genomes are stored. We are working on resolving this.

Helga

[igv-help]

For recent versions of IGV (> v2.16.0) this problem can be fixed for many genome assemblies by taking the following steps

(1) Set the "Genome server URL" in the advanced preferences (View > Preferences >Advanced) to  https://igv.org/genomes/genomes.tsv

(2) Refresh your genome from the menu "Genomes > Select hosted genome..."

If that does not fix the problem, and your IGV version is >= 2.16.0,  let us know.  

[Lillian Wu]

Hi,

I tried the above method on IGV 2.16.2, however the human CHM13-T2T genomes (v1.1 and v2.0) still cannot be loaded.

Thank you,

Lillian 

[James Robinson]

v1.1 will not be loadable until the bucket issue is resolved, however I can load CHMT13-T2T v2.0 in IGV 2.18.2.   

--

---
You received this message because you are subscribed to a topic in the Google Groups "igv-help" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/igv-help/1sjq1UUdyso/unsubscribe.
To unsubscribe from this group and all its topics, send an email to igv-help+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/igv-help/01014167-bd99-499f-ad6c-bfe3eeb0c9ccn%40googlegroups.com.

[Adam Clayton]

I think this error (or a similar one) has returned again today, though intermittent at the moment. According to the log, it looks like it's getting stuck trying to access https://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/ncbiRefSeq.txt.gz

Thanks for investigating (again).

-Adam

[James Robinson]

It's working for me,  could you try again?   

[Adam Clayton]

It's working for me again too. This is the second time today we've seen a hiccup in access to one or more resources lasting around 15-20 mins. Hopefully just a temporary glitch.

[James Robinson]

Yes hopefully, that's really out of our control.   The only way to 100% prevent this is to download the files and host them locally.    For the annotation files (e.g. ncbiRefSeq.txt.gz) this is actually easy to do.   The genome configuration is in the folder "<user home>/igv/genomes.  For hg19 this is hg19.json.   You can download the annotation file then edit hg19.json to point to your local copy.   For example, downloading ncbiRefSeq.txt.gz and storing it in your igv/genomes folder, the url (which can be a local file path)  would simply be

"url": "ncbiRefSeq.txt.gz"

The full hg19.json for this example is shown below.   This is more difficult to do for the sequence data as the files are large, but possible in principle.    BTW the fasta file and its index are only there to support old versions of IGV, no longer needed.

{

  "id": "hg19",

  "name": "Human (GRCh37/hg19)",

  "fastaURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg19/hg19.fasta",

  "indexURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg19/hg19.fasta.fai",

  "cytobandURL": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/cytoBand.txt.gz",

  "aliasURL": "https://igv.org/genomes/data/hg19/hg19_alias.tab",

  "twoBitURL": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.2bit",

  "chromSizesURL": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.chrom.sizes",

  "tracks": [

    {

      "name": "Refseq Genes",

      "format": "refgene",

      "id": "hg19_genes",

      "url": "ncbiRefSeq.txt.gz",

      "indexed": false,

      "order": 1000000,

      "infoURL": "https://www.ncbi.nlm.nih.gov/gene/?term=$$"

    }

  ],

  "chromosomeOrder": "chr1,chr2,chr3,chr4,chr5,chr6,chr7,chr8,chr9,chr10,chr11,chr12,chr13,chr14,chr15,chr16,chr17,chr18,chr19,chr20,chr21,chr22,chrX,chrY"

}

--

---
You received this message because you are subscribed to a topic in the Google Groups "igv-help" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/igv-help/1sjq1UUdyso/unsubscribe.
To unsubscribe from this group and all its topics, send an email to igv-help+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/igv-help/5cb97d74-6fd3-4f5e-97e1-a401394d0d7cn%40googlegroups.com.

[Adam Clayton]

Thanks for the info. If the issues become persistent, we'll explore setting up local annotations.

[Adam Clayton]

We're experiencing another Connection time out with IGV (now v2.18.2) unable to download hg19 or hg38 genome data into IGV. I tried to set up local copies, but waited to try and download and the data (until now), so I can't finish setting things up.

Anything else we can try in the meantime?

-Adam

[James Robinson]

Not if the UCSC download server is down.   It should not be down for long.

[Adam Clayton]

Ok servers are back up. I've downloaded all of the files using the URLs provided in the hg19.json above except for the ncbiRefSeq.txt.gz. Can you confirm the origin URL for that file (i.e. which version of this file is being loaded by IGV by default)? 

I found the goldenPath archive for hg19 and a few version listed there. https://hgdownload.soe.ucsc.edu/goldenPath/archive/hg19/ncbiRefSeq/

Just want to make sure I'm using the same version as IGV for the local data.

Thanks again!

[James Robinson]

The genome.json for hg19 is below.  I have removed the fastaURL and indexURL as you don't need them for recent versions of IGV.

{
  "id": "hg19",
  "name": "Human (GRCh37/hg19)",

  "cytobandURL": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/cytoBand.txt.gz",
  "aliasURL": "https://igv.org/genomes/data/hg19/hg19_alias.tab",
  "twoBitURL": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.2bit",
  "chromSizesURL": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.chrom.sizes",
  "tracks": [
    {
      "name": "Refseq Genes",
      "format": "refgene",
      "id": "hg19_genes",

      "url": "https://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/ncbiRefSeq.txt.gz",

      "indexed": false,
      "order": 1000000,
      "infoURL": "https://www.ncbi.nlm.nih.gov/gene/?term=$$"
    }
  ],
  "chromosomeOrder": "chr1,chr2,chr3,chr4,chr5,chr6,chr7,chr8,chr9,chr10,chr11,chr12,chr13,chr14,chr15,chr16,chr17,chr18,chr19,chr20,chr21,chr22,chrX,chrY"
}

--

---
You received this message because you are subscribed to a topic in the Google Groups "igv-help" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/igv-help/1sjq1UUdyso/unsubscribe.
To unsubscribe from this group and all its topics, send an email to igv-help+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/igv-help/cc5bd74b-0673-4579-a05b-e1c000b3d4d1n%40googlegroups.com.

[Adam Clayton]

Thank you James, I'm trying to load hg19 locally by downloading all of the files in the URLs of your json above and storing them in the genomes folder with the hg19.json file which I've modified as follows below. When I try to load the genome and choose the hg19.json, I'm getting the following error " hg19.2bit (The system cannot find the file specified)" even though the 2bit file is in the same directory. Sorry if I've misunderstood your previous explanation. Thanks for your continued support.

{
  "id": "hg19",
  "name": "Human (GRCh37/hg19)",

  "cytobandURL": "cytoBand.txt.gz",
  "aliasURL": "hg19_alias.tab",
  "twoBitURL": "hg19.2bit",
  "chromSizesURL": "hg19.chrom.sizes",

  "tracks": [
    {
      "name": "Refseq Genes",
      "format": "refgene",
      "id": "hg19_genes",

      "url": "ncbiRefSeq.txt.gz",

      "indexed": false,
      "order": 1000000,
      "infoURL": "https://www.ncbi.nlm.nih.gov/gene/?term=$$"
    }
  ],
  "chromosomeOrder": "chr1,chr2,chr3,chr4,chr5,chr6,chr7,chr8,chr9,chr10,chr11,chr12,chr13,chr14,chr15,chr16,chr17,chr18,chr19,chr20,chr21,chr22,chrX,chrY"
}

[James Robinson]

Apologies, it should work like that but it looks like there is a bug.   Could you try using absolute paths?   That should work.    Also, it's not necessary that those files be in the same folder as the "json" file, it's o.k. if they are just not necessary.

I will look into the relative path problem and fix it for the next release.

To view this discussion on the web visit https://groups.google.com/d/msgid/igv-help/0094f817-815b-4c71-948a-5a1824752229n%40googlegroups.com.

[igv-help]

If you would like to track progress I opened an issue for myself here: https://github.com/igvteam/igv/issues/1587