[Bioperl-l] Bio::Tools::BPbl2seq

[Peter Schattner]

Man-Suen Chan wrote:

> I have been using Bio::Tools::BPbl2seq to blast two sequences against each other where the query is a full length sequence and the subject is a shorter sequence which is repeated in the longer sequence. The bl2seq report shows several hits corresponding to these repeated regions. However the bioperl module appears to allow processing of the top hit only - (unless I have missed something).
>

I have finally gotten around to looking at this problem and there is indeed a bug in BPbl2seq.
.
I am in the process of fixing this bug.  This will necessitate modification of calling scripts so that obtaining, say, the score of the first HSP will change from:

 $report = new Bio::Tools::BPbl2seq(-file => Bio::Root::IO->catfile("t","bl2seq.out"));
 $score = $report->score;

to

$report = new Bio::Tools::BPbl2seq(-file => Bio::Root::IO->catfile("t","bl2seq.out"));
$hsp=  $report->next_feature;
$score = $hsp->score;

which is syntax already used in BPlite.

The reason I bring this up is that scripts using BPbl2seq written since the  0.7 release (which probably worked if there was only one HSP in the bl2seq report) will now break.   I intend to display of an error message if a script uses the now-unsupported syntax.  The alternative would appear to be to redefine the old methods calls internally so they still work - but this seems like a lot of
work to save probably a small number of scripts which will anyway work only on a limited number of bl2seq reports.

Any comments or objections to this approach?

Peter