[ensembl-dev] incorporation of dbSNP 132 into ensembl

Andrea Edwards edwardsa at cs.man.ac.uk
Fri Dec 10 21:14:01 GMT 2010


Thanks a lot

What about going the opposite way...

what code would you use to find genomic regions that map to more than 
one dbSNP id (or variation feature id).
It can be any genomic region, just for exploratory purposes.

On 10/12/2010 16:41, Pablo Marin-Garcia wrote:
>
>
> Hello Andrea,
>
> I forwarded your question to the ensembl dev mailing list because is 
> better to
> keep the discussions there. And you will have the chance of being 
> answered by
> the ensembl-variation people that are the ones with the most up to date
> knowledge of the API and data.
>
> My answer is inlinned:
>
> On Fri, 10 Dec 2010, Andrea Edwards wrote:
>
>> Hi Pablo
>>
>> Sorry to bother you again, I was hoping i could ask you a quick 
>> question.
>>
>> What happens to dbSNP records that are mapped to multiple locations. 
>> Lets say i had a fictitious record rs100 mapped to chromosome 10 at 
>> locui100bp and 200bp and i had this code
>>
>> $slice = $sa->fetch_by_region('chromosome', '10');
>> foreach $vf (@{$vfa->fetch_all_by_Slice($slice)}) {
>>
>> Would i get 2 distinct vf objects for the dbsnp record in this loop; 
>> one at 100bp and one at 200bp and both vf objs would have a 
>> map_weight of 2 (ignoring any other complexities)
>
> yes you would have the two vf. The concept of 'feature' means 
> something mapped, so if an object  maps several times in the genome it 
> would be stored as several different feature entries (one per mapping).
>
> This is way the method used to retrieve the variation-features from a 
> variation has the 'all' in it (fetch_all_by_variation) because this is 
> a 'one to many' relationship.
>
>
>   -Pablo
>>
>> thanks a lot
>>
>
>
> -----
>
>   Pablo Marin-Garcia
>
>
>





More information about the Dev mailing list