Hi Emmanuel,
thank for your suggestion. I do agree with you that what you describe would
be a better arrangement for the end user.

The IDs lists handling is currently delegated to the backend RDBMS. There
is also some batching processing logic to improve memory footprint on top
of that. These two factors determine the output order.
We can of course add some simple processing to maintain the input order
throughout this process but need to make sure that there will be no adverse
effect on speed and memory.


a





On Mon, Jan 7, 2013 at 11:12 PM, Emmanuel Levy <[email protected]>wrote:

> Hello,
>
> I sent an email to the Bioconductor mailing list but was redirected
> here for the following suggestion:
>
> It'd be (even more) convenient if Biomart would return data in an
> order identical to that given as input.
>
> E.g., if retrieving cDNA of IDs 1 to 10, this means sequences would be
> returned in the order 1 to 10 (i.e., not randomly).
>
> In other words, the First In would be the First Out (FIFO).
>
> Please find below a practical example of what I mean:
>
> > ensembl.yeast = useMart(biomart="ensembl", dataset =
> "scerevisiae_gene_ensembl")
> > ids = c("S000005528","S000000221" ,"S000005993", "S000003105",
> "S000004200")
> > yeast.gen = getSequence( id=ids, mart=ensembl.yeast, seqType=c("cdna"),
> type="sgd_transcript" )
> > yeast.gen[[2]]
> [1] "S000005528" "S000004200" "S000003105" "S000000221" "S000005993"
>
> All the best,
>
> Emmanuel
> _______________________________________________
> Users mailing list
> [email protected]
> https://lists.biomart.org/mailman/listinfo/users
>
_______________________________________________
Users mailing list
[email protected]
https://lists.biomart.org/mailman/listinfo/users

Reply via email to