On 29/09/08 11:38 AM, "Dan Staines" <[EMAIL PROTECTED]> wrote:
>
Hi Dan,
The dumps would be most helpful for us to reproduce the problem over here
Thank you,
A.
>
>
> Victor Gu wrote:
>> Hi, Dan,
>>
>> I am under the impression that the problem in your email is related
>> to the SQL generation problem when creating Mart from a multi-species
>> partition using MartBuilder. Could you outline the details steps how
>> the problem was produced?
>
> I think this is a different problem. In this instance, I'm producing a
> single mart for each set of species in a single ensembl database, rather
> than trying to do any further partitioning of the dataset.
>
> To summarise, this is using a pretty standard ensembl mart definition,
> based on the one used for core ensembl 50. We have used this
> successfully for both single and multi-species databases in the past,
> using an empty compara schema. However, we now have a populated compara
> schema containing data for one of the sets of species (i.e. for a single
> ensembl schema), and while we can readily build a mart, it doesn't
> include any homology tables.
>
> I can probably make some test dumps available if you need me to.
>
>> By the way, I have not had the time looking into the cause of the SQL
>> generation problem in a multi-species partition, shomething hopefully
>> I will be able to tackle soon.
>
> Not a problem - we ended up writing some scripts to slice up the
> generated mart into individual species instead.
>
> --
> Dan Staines, PhD
> EMBL-European Bioinformatics Institute Tel: +44-(0)1223-492507
> Wellcome Trust Genome Campus, Hinxton Fax: +44-(0)1223-494468
> Cambridge CB10 1SD, UK email: [EMAIL PROTECTED]
