[ensembl-dev] VEP dbNSFP error

Tue Jun 4 13:15:50 BST 2019

Dear Abhishek,

could you please replace the following line in the dbNSFP plugin: https://github.com/Ensembl/VEP_plugins/blob/release/93/dbNSFP.pm#L122 <https://github.com/Ensembl/VEP_plugins/blob/release/93/dbNSFP.pm#L122>
with: open HEAD, "tabix -fh $file 1:1-1 | “

I think we are throwing away potential error messages returned by the tabix program. Please run your vep command again after the edit and let me know if we you get any additional error messages.

Thank you,
Anja

> On 4 Jun 2019, at 11:00, Abhishek Nag <abhishek at well.ox.ac.uk> wrote:
> 
> Dear Anja,
> 
> Please see below my responses to your queries in bold. Hope that helps.
> 
> Thanks
> Abhishek
> From: Dev [dev-bounces at ensembl.org <mailto:dev-bounces at ensembl.org>] on behalf of Anja Thormann [anja at ebi.ac.uk <mailto:anja at ebi.ac.uk>]
> Sent: Monday, June 03, 2019 3:53 PM
> To: Ensembl developers list
> Subject: Re: [ensembl-dev] VEP dbNSFP error
> 
> Thank you for the extra details.
> 
> Could you please confirm one more time the return value for running:
> 
> tabix -fh /home/VEP/Plugins/dbNSFP4.0b2a.txt.gz 1:1-1 2>&1
> 
> It should be #chr    pos(1-based)    ref     alt     aaref   aaalt   rs_dbSNP151     hg19_chr ….
> 
> I get exactly the same output as you mentioned, on running the tabix -fh command
> 
> The error (ERROR: Could not read headers from dbNSFP4.0b2a.txt.gz) most likely points to a missing header line in the dbNSFP file.
> 
> If the plugin couldn't read from the dbNSFP file the error would be ERROR: Data file dbNSFP4.0b2a.txt.gz not found
> 
> There are 2 things which slightly confuse me:
> 
> 1) If you run vep with /apps/well/ensembl-tools/93/ensembl-vep/vep -i variants.txt -o output.txt --cache --dir /home/VEP/ --plugin dbNSFP, /home/VEP/Plugins/dbNSFP4.0b2a.txt.gz
> ALL --tab --force_overwrite --port 3337
> I would expect the full path name in the error message: ERROR: Could not read headers from /home/VEP/Plugins/dbNSFP4.0b2a.txt.gz
> 
> Yes, true, I do get the full path name in the error message
> 
> 2) tabix -H is not supported by tabix version: 0.2.5. But you are able to run tabix -H?
> 
> When I just run tabix (version 0.2.5) from my command line, the list of available options shows that the -H option for printing just the header line is available.
> 
> Once we figured out what is going on you need to switch to a more recent release version for the plugin <https://github.com/Ensembl/VEP_plugins/blob/release/97/dbNSFP.pm>. In the more recent versions of the dbNSFP plugin we are dealing with the different header names for grch37 and grch38 coordinates that have been introduced by dbNSFP for their last file version 2.*, 3.* and 4.*. But this is only relevant after the plugin has retrieved the header line.
> 
> Thank you,
> Anja
> 
> 
>> On 3 Jun 2019, at 14:43, Abhishek Nag <abhishek at well.ox.ac.uk <mailto:abhishek at well.ox.ac.uk>> wrote:
>> 
>> Dear Anja,
>> 
>> Thanks for your response.
>> 
>> I'm using the VEP version 93 and perl version v5.16.3
>> 
>> Thanks
>> Abhishek
>> From: Dev [dev-bounces at ensembl.org <mailto:dev-bounces at ensembl.org>] on behalf of Anja Thormann [anja at ebi.ac.uk <mailto:anja at ebi.ac.uk>]
>> Sent: Monday, June 03, 2019 2:25 PM
>> To: Ensembl developers list
>> Subject: Re: [ensembl-dev] VEP dbNSFP error
>> 
>> Dear Abhishek,
>> 
>> so far I wasn’t able to reproduce the error. I have been running tests with the same dbNSFP file and tabix version. Could you please also let me know your perl version and your vep version?
>> 
>> Best regards,
>> Anja
>> 
>>> On 26 May 2019, at 15:47, Abhishek Nag <abhishek at well.ox.ac.uk <mailto:abhishek at well.ox.ac.uk>> wrote:
>>> 
>>> Dear Anja,
>>> 
>>> Thanks for responding to my email. 
>>> 
>>> The header line that is returned on running tabix -H seems fine. 
>>> 
>>> Here is the VEP command that I'm running:
>>> 
>>> /apps/well/ensembl-tools/93/ensembl-vep/vep -i variants.txt -o output.txt --cache --dir /home/VEP/ --plugin dbNSFP,/home/VEP/Plugins/dbNSFP4.0b2a.txt.gz,ALL --tab --force_overwrite --port 3337
>>> 
>>> 
>>> 
>>> The /home/VEP/ path contains the Plugins subdirectory (which contains the dbNSFP4.0b2a.txt.gz file) as well as the cache folder homo_sapiens.
>>> 
>>> 
>>> 
>>> Thanks for your help!
>>> 
>>> 
>>> 
>>> Abhishek
>>> 
>>> 
>>> From: Dev [dev-bounces at ensembl.org <mailto:dev-bounces at ensembl.org>] on behalf of Anja Thormann [anja at ebi.ac.uk <mailto:anja at ebi.ac.uk>]
>>> Sent: Thursday, May 23, 2019 3:52 PM
>>> To: Ensembl developers list
>>> Subject: Re: [ensembl-dev] VEP dbNSFP error
>>> 
>>> Dear Abhishek,
>>> 
>>> could you please run tabix -H dbNSFP4.0b2a.txt.gz and let me know if it returns the header line starting with #chr  pos(1-based)..? If that is the case could you please let me know who exactly your are running the VEP command together with the dbNSFP plugin command? If the header is returned correctly I’m wondering if the dbNSFP file path is maybe not found correctly by the plugin.
>>> 
>>> Regards,
>>> Anja
>>> 
>>>> On 23 May 2019, at 13:53, Abhishek Nag <abhishek at well.ox.ac.uk <mailto:abhishek at well.ox.ac.uk>> wrote:
>>>> 
>>>> Hi,
>>>> 
>>>> I have been trying to use the dbNSFP database to annotate my variants. Currently, I'm trying to use the latest version of the database (v4.0), but I have also tried using the version 3.5. With both versions, I'm getting the following error message whenever I'm trying to run it -
>>>> 
>>>> Failed to instantiate plugin dbNSFP: ERROR: Could not read headers from dbNSFP4.0b2a.txt.gz
>>>> 
>>>> 
>>>> 
>>>> When I zcat the database file, I am able to read the header line, which was obtained from one of the chromosome files, as suggested in the online manual.
>>>> 
>>>> 
>>>> 
>>>> The formatting steps for the database (including tab indexing) seem to have gone fine, so I'm not sure where the problem is. Also, I am being able to use the other VEP Plugins just fine; for now, I'm getting this error messge with only the dbNSFP plugin.
>>>> 
>>>> 
>>>> 
>>>> It would be great if I could have some pointers regarding where the issue might be.  
>>>> 
>>>> 
>>>> Many Thanks
>>>> Abhishek
>>>> _______________________________________________
>>>> Dev mailing list    Dev at ensembl.org <mailto:Dev at ensembl.org>
>>>> Posting guidelines and subscribe/unsubscribe info: https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org <https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org>
>>>> Ensembl Blog: http://www.ensembl.info/ <http://www.ensembl.info/>
>>> _______________________________________________
>>> Dev mailing list    Dev at ensembl.org <mailto:Dev at ensembl.org>
>>> Posting guidelines and subscribe/unsubscribe info: https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org <https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org>
>>> Ensembl Blog: http://www.ensembl.info/ <http://www.ensembl.info/>
>> _______________________________________________
>> Dev mailing list    Dev at ensembl.org <mailto:Dev at ensembl.org>
>> Posting guidelines and subscribe/unsubscribe info: https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org <https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org>
>> Ensembl Blog: http://www.ensembl.info/ <http://www.ensembl.info/>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org <mailto:Dev at ensembl.org>
> Posting guidelines and subscribe/unsubscribe info: https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org <https://lists.ensembl.org/mailman/listinfo/dev_ensembl.org>
> Ensembl Blog: http://www.ensembl.info/ <http://www.ensembl.info/>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20190604/2b56d0cf/attachment.html>