[Bioperl-l] trying to save blast hit sequences to fasta file

Alicia Amadoz Alicia.Amadoz at uv.es
Thu Aug 2 03:06:54 EDT 2007


Hi, thanks for your help and suggestions. I have tried the example code
of Jay Hannah and it works perfectly. But what I need to save in fasta
format is the whole sequence in the database that is similar to my query
sequence. I don't understand very well the difference between
hit_string() and query_string(), are they the whole sequence that is
similiar (about hit_string), a part of the whole sequence or just the
part that is aligned to my query string? 

With the previous code what I have are different sequences in length
with the same id as my query string, so I am not sure that I am doing
what I need to do. Any light on this point?

Thank you very much for your help.
Alicia

> Alicia,
> 
> > Hi, I would like to save my hit sequences from a blast result in a fasta
> > file. I am trying some things but I have problems using Bio::SearchIO
> > and Bio::SeqIO. Hope anyone could help me with this. Here is my current
> > code:
> > # my $seq_out = Bio::SeqIO->new("-file" => ">$fasfilename", "-format" =>
> > "fasta");
> > my $seq_out = Bio::SearchIO->new("-file" => ">$fasfilename", "-format"
> > => "fasta");
> > ...
> >        my $hseq = $hsp->hit_string();
> >          # $seq_out->write_seq($hseq);
> >          $seq_out->write_result($hseq);
> 
> You have encountered two common problems for BioPerl beginners:
> 
> 1. "fasta" means two different things! In SearchIO it refers to the
> output format of the "fasta" sequence alignment software. In SeqIO it
> refers to a file format that stores just sequences. Confusing, I know.
> You need SeqIO and write_seq, not SearchIO and write_result.
> 
> 2. $hseq is a STRING which has the raw sequence letters in it.
> However, the write_seq() method needs a Bio::Seq object (which has
> extra details like the name and ID) not a raw string.
> 
> The example code Jay Hannah supplied in his reply looks pretty good,
> you should try it.
> 
> -- 
> --Torsten Seemann
> --Victorian Bioinformatics Consortium, Monash University
> 
> 




More information about the Bioperl-l mailing list