[Bioperl-l] extract feature seq when split between 2 GenBank accessions

[Charles Hauser]

Jason,

Thanks.

On Wed, 2003-08-27 at 13:05, Jason Stajich wrote:
> If you are getting the seq via spliced_seq you can pass in a
> Bio::DB::RandomAccessI (either a [local] Bio::Index::Fasta or [remote]
> Bio::DB::GenBank, etc db handle) to the spliced_seq object.
> 
> Now I think there is a bug because spliced seq is sorting the locations
> before processing on them which has been reported but not fixed
> (I am really hoping for some more bugfixing developers out there folks!)
> but it should work through that system once that bug is fixed.

I'm using spliced_seq, and its returning sequence w/ N's for the
segments derived form joins outside the current accession.

> 
> I would just use a Bio::DB::Fasta/Bio::Index::Fasta where you have the
> accessions indexed instead of reading in all the possible seqs and storing
> in a hash to keep the memory requirements down.  You can also use the
> DB::Failover + DB::FileCache to cache local/remote calls if you need to
> mix local and remote dbs.

I'll try this -thanks.

Charles