[Bioperl-l] wrong nomatch position from protein with singly deleted Aa

[Jason Stajich]

I know you are trying to give an example but this isn't really enough  
for someone to help as you are referring to a sequence alignment we  
can't see.  I can't tell if this a problem with translated blast  
coordinates, the seq_inds code alone, or what.

Why don't you gather a sample report, the code you are using, and  
your expected result together in something that someone can run and  
submit it as a bug to bugzilla.  Then hopefully someone from the  
community will download and reproduce your problem for you and can  
tell whether the problem is in the module or elsewhere.

-jason

On Jan 21, 2006, at 1:40 PM, Anders Stegmann wrote:

> Hi BioPerl!
>
> I have an original protein seq which I blastp (standalone) against  
> the same seq with Aa nr 61 deleted manually.
>
> The result is that the subject nomatch is Aa. E on position 60,  
> which is definitely not a mismatch!!?
> This also happens if I delete two Aa at positions 61 and 62 in the  
> subject seq.
> This does strangely enough not happen if I delete a whole line (60  
> Aa) in the subject seq.
>
> The result for the query  nomatch is Aa. V at position 61, which is  
> korrekt (the subrutine code is similar to the subject code shown  
> below).
>
>
> the code I use is following:
>
> sub subject_seq_alignment_nomatch_residues {
>
> my ($hsp_obj) = @_;
> my %subject_nomatch_hash = ();
> my @new_subject_string = ();
>
> my @subject_string = split , $$hsp_obj->hit_string;
>
> foreach (@subject_string) { #positioner i visse tilfælde
>
> if ($_ ne '-') {push @new_subject_string, $_};
>
> }
>
> my $start_subject_number = $$hsp_obj->start('hit');
>
> $start_subject_number = $start_subject_number - 1;
>
> foreach ($$hsp_obj->seq_inds('hit', 'nomatch')) {
>
> $subject_nomatch_hash{$_} = $new_subject_string[$_ -1 - 
> $start_subject_number];#positionen, trækker derefter den  
> tilsvarende #aminosyre ud af subjekt sekvensen
> }
>
> return %subject_nomatch_hash;
>
> }
>
> It has nothing to do with the foreach (@subject_string) { code or  
> the $start_subject_number (cause it is 0 in this example). I checked!
>
> How can this be?
>
> Regards Anders.
>
>
>
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at portal.open-bio.org
> http://portal.open-bio.org/mailman/listinfo/bioperl-l

--
Jason Stajich
Duke University
http://www.duke.edu/~jes12/