1 Feb 20:03
Odd issue with HSP::seq_inds()
Chris Fields <cjfields <at> uiuc.edu>
2008-02-01 19:03:36 GMT
2008-02-01 19:03:36 GMT
Apologies ahead of time for the longish post! While working on a recent bug (http://bugzilla.open-bio.org/show_bug.cgi?id=2436 ) I found a pretty significant bug. It appears that seq_inds() is giving some bad return values tied specifically to reports where either or both query or hit are translated. I'll file a full bug report but I felt it might be worth bringing up here to discuss specifically what values should be expected when these reports are parsed (all three nucleotide positions or only the first for each nonmatching translated position, etc). I used the following script to track the total HSPs as well as those where seq_inds() returns positions within the HSP bounds for query and hit: ------------------------------------------- use Bio::SearchIO; use Test::More qw(no_plan); my $file = shift || die; my $io = Bio::SearchIO->new(-format => 'blast', -file => $file); my ($hsp_co, $correctq, $correcth) = (0,0,0); while (my $res = $io->next_result) { while (my $hit = $res->next_hit) { while (my $hsp = $hit->next_hsp) {(Continue reading)
RSS Feed