[ensembl-dev] Missing MAF numbers on VEP REST service

Paul Flicek flicek at ebi.ac.uk
Mon Aug 1 08:19:02 BST 2016


All interfaces should return always return the same information when 
asked the same question.  When this does not happen, we consider this a 
bug and will look into it.

Regarding what should be returned, it sounds like you are suggesting 
that we return (1 - Major allele frequency) and that we sum the 
frequencies of all of the various non-major alleles.  Is this what you 
mean?

Would it be better to return the frequency of all of the alleles and 
have this calculation be done explicitly by the requestor?


Thanks,
Paul	


On 31 Jul 2016, at 15:50, Wolf Beat wrote:

> Excellent question.
>
> I essentially expect the REST interface to return the same information 
> as the website and the biomart API.
> Now, it could be that biomart and the website are both wrong in 
> displaying the MAF number they display, but it should be coherent 
> between all three ways to access.
>
> What i would expect personnaly is:
> As the 3 alleles are regrouped under the same rs number, i need to 
> have the MAF of all 3 alleles combined. That what i would expect, as 
> the rs number indicates that all 3 alleles map to it. I think thats a 
> reasonably assumption?
>
> Kind regards
>
> Beat Wolf
> ________________________________________
> From: dev-bounces at ensembl.org [dev-bounces at ensembl.org] on behalf of 
> Paul Flicek [flicek at ebi.ac.uk]
> Sent: Sunday, July 31, 2016 3:38 PM
> To: Ensembl developers list
> Subject: Re: [ensembl-dev] Missing MAF numbers on VEP REST service
>
> Hi Beat,
>
> Could you give me some insight about what you are looking for in the
> case of SNPs with more than two alleles.  For example, what do you
> expect the MAF to return in these cases?
>
> More specifically, MAF is not a property of a variant site itself.
> Frequency is a property of the various alleles.  We are currently
> loading and returning the dbSNP MAF for each SNP, which is defined as
> the frequency of the second most common allele, but in reality, the 
> SNP
> has several alleles and each of them has a frequency.  As more
> individuals are assayed, more SNPs will be found to have 3 or 4 
> alleles,
> so this is an issue that will grow.
>
> Thanks,
> Paul
>
>
>
> On 31 Jul 2016, at 8:39, Wolf Beat wrote:
>
>> Hello,
>>
>> i recently migrated a lot of my code from the biomart interface to 
>> the
>> ensembl rest service. I'm mainly using grch37 as my source.
>> I made the following observation:
>> When i query a multi allelic variant, the REST VEP service returns 
>> not
>> MAF numbers.
>> But when i query the biomart service or check the variant on the
>> website manually, the MAF numbers are present.
>> Just to give one example:
>> Variant rs6687605 has an MAF number of 0.43 when looked at on the
>> website and biomart:
>> http://grch37.ensembl.org/Homo_sapiens/Variation/Explore?r=1:25889132-25890132;v=rs6687605;vdb=variation;vf=106422589
>>
>> But the REST service returns no MAF number:
>> http://grch37.rest.ensembl.org/vep/hsapiens/id/rs6687605?content-type=application/json
>>
>> I observed the same behaviour for:
>> rs2229546, rs6022, rs1058111
>>
>> I also observed that the VEP rest service specifies:
>> http://grch37.rest.ensembl.org/documentation/info/vep_id_get
>> That the ExAC information is by default off. But i do recieve it
>> without specifying the option.
>> In my case i actually want it, but its still interesting to note.
>>
>> The first problem is kind of a big issue for me, any help in solving
>> it would be greatly appreciated.
>>
>> Kind regards
>>
>> Beat Wolf
>>
>> _______________________________________________
>> Dev mailing list    Dev at ensembl.org
>> Posting guidelines and subscribe/unsubscribe info:
>> http://lists.ensembl.org/mailman/listinfo/dev
>> Ensembl Blog: http://www.ensembl.info/
>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org
> Posting guidelines and subscribe/unsubscribe info: 
> http://lists.ensembl.org/mailman/listinfo/dev
> Ensembl Blog: http://www.ensembl.info/
>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org
> Posting guidelines and subscribe/unsubscribe info: 
> http://lists.ensembl.org/mailman/listinfo/dev
> Ensembl Blog: http://www.ensembl.info/




More information about the Dev mailing list