HGVSc column blank or error

[Miu ki Yip]

Hello,

 

Would it be possible to explain how the HGVSc column is populated? We are having an issue where a cohort being loaded in does one of three things:

1. Shows the HGVSc

2. Is blank.

3. Shows "Error" (Unable to load data)

 

This github issue (https://github.com/cBioPortal/cbioportal/issues/6867) implies that this column is pulling mutation assessor, is this correct? How are some mutations able to have this column populated but not others (blanks and error)?

 

Please advise, thank you.

[Ino de Bruijn]

Hi Miu ki Yip,

Thanks for reaching out! The HGVSc data is pulled from Genome Nexus. The cBioPortal frontend sends over the mutation event (chrom,pos,ref,alt) and gets the hgvsc (and other annotations) back from genome nexus.

You can see the exact data it sends over in the network tab of the chrome browser debug tools, see e.g. for this Patient https://www.cbioportal.org/patient?studyId=lgg_ucsf_2014&caseId=P04:

Are you able to share some of the variants that respond in a blank or error result?

Best wishes,

Ino

--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cbioportal/BL0PR06MB465772E12B88A1279B280110ED4D9%40BL0PR06MB4657.namprd06.prod.outlook.com.

[Miu ki Yip]

Hi Ino,

 

Thanks for getting back to us. This is very helpful information.

 

As it turns out, some of the missing annotations was due to the chrom positioning being off by 1 after looking at the Genome Nexus page.

 

Cheers,

Miu ki

 

From: Ino de Bruijn <i...@ino.pm>
Sent: Thursday, April 15, 2021 11:38 AM
To: Miu ki Yip <miy...@med.cornell.edu>
Cc: cbiop...@googlegroups.com; Joel Oakley <joo...@med.cornell.edu>; Evan Fernandez <emf...@med.cornell.edu>
Subject: [EXTERNAL] Re: [cbioportal] HGVSc column blank or error

 

Hi Miu ki Yip,

 

Thanks for reaching out! The HGVSc data is pulled from Genome Nexus. The cBioPortal frontend sends over the mutation event (chrom,pos,ref,alt) and gets the hgvsc (and other annotations) back from genome nexus.

 

You can see the exact data it sends over in the network tab of the chrome browser debug tools, see e.g. for this Patient https://www.cbioportal.org/patient?studyId=lgg_ucsf_2014&caseId=P04:

 

 

Are you able to share some of the variants that respond in a blank or error result?

 

Best wishes,

Ino

 

[Miu ki Yip]

Hi Ino,

 

As a follow up, I see that for some samples with more than one mutation, if one of the mutations is N/A on Genome Nexus, all of the mutation are showing Error under Functional Impact and HGVSc in the same table. Is it possible to show the information for the mutations that have information even if one mutation is not available on Genome Nexus?

 

Thanks,

Miuki

[Ino de Bruijn]

Hi Miuki,

Thanks for letting us know! That sounds like a bug. It should indeed work for the other variants when only one of them is failing. We will investigate this issue further

What version of cbioportal are you using locally? Or do you see this issue in the public website (cbioportal.org) as well?

Thanks,

Ino

[Miu ki Yip]

Hi Ino,

 

We are using version 3.4.17. Has this been fixed in a newer release?

 

Thanks,

Miuki

[Ino de Bruijn]

Hi Miuki,

I imagine it is probably on the genome nexus server side, so maybe not related to your cbioportal version. One way to check would be if you could send us the curl request that is failing. If you want to keep the data private and not send it in the google group you can send it to my institutional email (debr...@mskcc.org). You can get the curl command by right clicking on the request in the network tab and selecting "copy as curl". You can also change the curl request a bit so that it only includes the one failing variant and maybe one other that should be successful. Then we can try and fix the genome nexus server on our end

Thanks,

Ino 

To view this discussion on the web visit https://groups.google.com/d/msgid/cbioportal/BL0PR06MB4657E5661F501F2C2B5F7121ED499%40BL0PR06MB4657.namprd06.prod.outlook.com.