Bioperl or ncbi: parsing refseq files

roibrodo has asked for the wisdom of the Perl Monks concerning the following question:

Hello,

I have downloaded some large RefSeq files (Genbank format, i.e. both sequence and annotations; e.g. http://www.ncbi.nlm.nih.gov/nuccore/NC_002945.

I want to get a part of that sequence along with its annotations - exactly as I do online using the "Change region shown" window on the right - then dump this as a (smaller) Genbank file. Very simple... but I couldn't figure out how to do that.

Thanks!

UPDATE: RESOLVED

one way of achieving this is using seqret from emboss: e.g. seqret -sformat1 gb -sequence "in_seq.genbank" -sbegin 10000 -send 50000 -feature -outseq out_seq.genbank.genbank -osformat2 gb

Thanks to vinnana from the bioperl community!

Comment on Bioperl or ncbi: parsing refseq files Select or Download Code

Replies are listed 'Best First'.
Re: Bioperl or ncbi: parsing refseq files by BioLion (Curate) on Jun 14, 2010 at 15:45 UTC
Hi, your flat genbank files can be handled using the BioPerl suite. See this HowTo for a very detailed guide to IO, which includes genbank parsers (and a lot of others). You'll be using the Bio::SeqIO module ( See Yes, even you can use CPAN for a guide on getting modules installed, or ignore if i am patronising you... ) to read in the files, test each sequence feature if it is in your region of interest, and if it is, write it out to a fresh (smaller) file. You can do all this on-the-fly, so your large file shouldn't trouble memory problems... Have a read, have a go, and get back to the Monastery (with examples and code) if you are having problems. Hope this helps! Update: Typos... Just a something something...	[reply]
Re^2: Bioperl or ncbi: parsing refseq files by roibrodo (Sexton) on Jun 15, 2010 at 09:30 UTC
Thanks for the reply. I'm not sure I got it right. Even if the feature is not fully in the region of interest, but only partially in it, I want to "truncate" it and take it. I also want the sequence (that appears after all the features) to be outputted. Basically, I want to do exactly what the "change region shown" does on the online version of NCBI. I would appreciate a more verbose example, if possible, since this are my first steps with BioPerl. Thanks!	[reply]
Re: Bioperl or ncbi: parsing refseq files by david_lyon (Sexton) on Jun 14, 2010 at 18:33 UTC
If I was doing this which I have done a million times you want to use one if the emboss packages which can parse out the features and sequences that you are interested in or spit out the features in a table format so you can upload them into a db or parse them on the fly. eg: http://emboss.sourceforge.net/apps/release/6.2/emboss/apps/extractfeat.html I would avoid using bioperl for this.	[reply]
Re^2: Bioperl or ncbi: parsing refseq files by roibrodo (Sexton) on Jun 15, 2010 at 09:36 UTC
Thanks for the reply. I read some of the documentation but did not find an appropriate example. Would you be kind enough to let some hints on how to get from a GenBank file to a smaller GenBank file containing some range of the sequence? It will really help getting stated with emboss (I think examples are a great way to start experimenting) Thanks!	[reply]