[Bioperl-l] Re: RegSeq and NT_****** contig Id

Jason Stajich jason at cgt.duhs.duke.edu
Mon Jul 14 21:54:48 EDT 2003


Also - with the changes to add gb| to the id the RefSeq tests don't work
on the main trunk - tried to fix but won't have time till next week

-j

On Mon, 14 Jul 2003, Heikki Lehvaslaiho wrote:

> Jing,
>
> Bio::DB::RefSeq used to inherit from Bio::DB::NCBIHelper, but lately it
> has been a subclass of Bio::DB::DBFetch. Looks like in the transition we
> lost the warning:
>
>   $self->throw("NT_ contigs are whole chromosome files which are
>     not part of regular database distributions. Go to
>     ftp://ftp.ncbi.nih.gov/genomes/.")
> 	if $ids =~ /NT_/;
>
> It also true that the NCBI Entrez web interface now allows retrieving
> NT_ contigs, so it would be possible to hack RefSeq class to retrieve
> them. However, NCBI has asked us help to limit the load to their online
> services, I am hesitant to do that when their eutils server is excluding
> them (Or is it? Do we just need different parameters?). Downloading a
> 28,477,090 base mouse chromosome 1 sequence with tons of annotation is
> certainly heavy. The warning should definitely be put back in.
>
> Yours,
> 	-Heikki
>
> P.S. DBI is for accessing local relational database and not needed here.
> 	-H.
>
> On Mon, 2003-07-14 at 19:36, jzhao wrote:
> > Dear Sir,
> >
> > I was trying to retrieve some mouse contig data from the RefSeq database
> > with Bioperl. My testing perl script looks like:
> >
> > use Bio::DB::RefSeq;
> > use Bio::SeqIO;
> > use DBI;
> > use strict;
> >
> > my $gb = new Bio::DB::RefSeq;
> > my $seq = $gb->get_Seq_by_acc('NT_039167');
> >
> > if ( defined $seq ) {
> > 	printf 'seq defined', '\n';
> > }
> > else {
> > 	printf 'seq undefined', '\n';
> > }
> >
> > This script works with access ids like NC_000913 (bacteria genome), but
> > with NT_****** contig id, the $seq returns undefined. I checked, these
> > contig data are stored in RefSeq db ftp site, but why they are not
> > available through DBI interface? anything I'm missing here?
> >
> > Thank you very much,
> > Jing
>

--
Jason Stajich
Duke University
jason at cgt.mc.duke.edu


More information about the Bioperl-l mailing list