Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-09-24

| Channels | #bioperl index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:21 kyanardag_ joined #bioperl
01:26 kyanardag_ left #bioperl
01:39 kyanardag_ joined #bioperl
02:42 kyanardag_ left #bioperl
03:35 kyanardag_ joined #bioperl
03:41 kyanardag_ left #bioperl
03:54 kyanardag_ joined #bioperl
04:16 dnewkirk left #bioperl
06:04 deafferret left #bioperl
07:11 ank left #bioperl
07:33 kyanardag_ left #bioperl
07:35 kyanardag_ joined #bioperl
11:17 carandraug joined #bioperl
11:21 brandi1 joined #bioperl
12:09 brandi1 left #bioperl
13:43 brandi1 joined #bioperl
13:44 brandi1 left #bioperl
14:27 randomubuntuguy joined #bioperl
14:32 dnewkirk joined #bioperl
14:45 ank joined #bioperl
15:04 JunY left #bioperl
15:17 randomubuntuguy left #bioperl
15:17 jhannah joined #bioperl
15:18 jhannah is now known as deafferret
15:19 kyanardag_ left #bioperl
15:36 deafferret left #bioperl
15:40 jhannah joined #bioperl
15:41 jhannah left #bioperl
15:42 jhannah joined #bioperl
15:50 kyanardag joined #bioperl
16:15 pittpgc joined #bioperl
16:15 pittpgc hello
16:15 pittpgc anybody know about Clustalw and SimpleAlign?
16:16 pittpgc here is my problem....
16:17 pittpgc when I use Bio::Tools::Run::Alignment::Clustalw to generate an alignment, it uses the "." (period) character to represent gaps
16:17 jhannah are you on bioperl-live?
16:17 pittpgc yes
16:17 jhannah phew. step 1 looks good  :)
16:17 pittpgc $aln->remove_gaps(".") does not work
16:18 pittpgc it removes all columns
16:18 jhannah I've never tried remove_gaps(). Are there tests for it in t/ ?
16:18 pittpgc don't know
16:18 pittpgc i've used remove_gaps successfully in the past but with "-"
16:19 jhannah out of curiosity, have you tried a s/\./-/g on your input data and then seen if remove_gaps suddenly works?
16:20 pittpgc i haven't tried that but I can almost assure you that it will work
16:20 arcsine that
16:20 pittpgc $gap_line .= (exists $gap_hsh{$pos}) ? $gapchar:'.';     <- sub gap_line in SimpleAlign
16:20 pittpgc line # 1678
16:20 arcsine may be easier than trying to
16:20 jhannah oops... lunch appt. biab
16:20 arcsine get this to work using '.'
16:21 pittpgc yeah gap_line uses "." as a special character
16:21 pittpgc and does not handle the case where the gap_char could also be equal to "."
16:21 arcsine right
16:22 pittpgc see the problem is that Bio::Tools::Run::Alignment::Clustalw returns gaps as "."
16:22 pittpgc so how can I in my SimpleAlign object change all values of "." to "-"?
16:24 arcsine does gap_char work as expected?
16:24 arcsine i think you set the gap character using it
16:24 pittpgc yes, gap_char will set the gapchar to "."
16:24 pittpgc but that's not the problem
16:25 pittpgc gap_line returns a string of all "." so therefore all columns are removed from the alignment
16:25 arcsine ....i wonder
16:25 arcsine if you escaped it
16:25 arcsine if it would work
16:26 pittpgc Line 1958 $self->throw("Single gap character, not [$char]!")      <- gap_char doesn't accept multiple character string
16:27 arcsine damn
16:29 pittpgc problem is that the Bioperl clustalw module is returning an alignment with "." as the gap character
16:29 pittpgc when you run Clustalw from the command line (no Bioperl), it uses "-
16:29 pittpgc "-" as the gap char
16:29 pittpgc so somewhere, bioperl is converting the "-" to "."
16:36 arcsine what you may be able to do
16:36 arcsine to solve your immediate issue
16:37 arcsine is use AlignIO to dump out the alignment
16:37 arcsine then change the character
16:37 arcsine and read it back in
16:41 pittpgc yeah I guess so....seems clunky though
16:41 arcsine totally
16:41 arcsine what version of clustalw you using
16:42 pittpgc maybe there needs to be a method in SimpleAlign to change the gap character in the actual sequences in the alignment object to whatever you want
16:42 pittpgc version 2.0.12
16:42 pittpgc shoot gotta run
16:42 pittpgc thanks for your help
16:43 arcsine laters
16:43 arcsine fyi:
16:43 arcsine Clustalw.pm has
16:43 arcsine only been tested using version 1.8 of clustalw.
16:44 arcsine also:
16:44 arcsine Optionally, the
16:44 arcsine factory may be passed most of the parameters or switches of the
16:44 arcsine clustalw program, e.g.:
16:44 arcsine @params
16:44 arcsine = ('ktuple' => 2, 'matrix' => 'BLOSUM');
16:44 arcsine $factory = Bio::Tools::Run::Alignment​::Clustalw->new(@params);
16:53 philsf joined #bioperl
16:53 philsf left #bioperl
16:53 philsf joined #bioperl
17:19 randomubuntuguy joined #bioperl
17:53 randomubuntuguy left #bioperl
19:03 vinnana left #bioperl
19:20 pittpgc is there code somewhere that converts one alignment format to another (i.e. clustal => phylip)
19:29 jhannah Bio::AlignIO does exactly that
19:29 jhannah pittpgc: ^
19:30 jhannah is now known as deafferret
19:41 pyrimidine joined #bioperl
19:43 pyrimidine arcsine: ClustalW v2 should work with BioPerl now; I had updated that at some point (I think  prior to last CPAN release, 1.6.1)
19:50 pittpgc left #bioperl
19:54 philsf left #bioperl
20:32 arcsine good to know
20:33 arcsine i'm really new to bioperl, but it has been really useful so far
20:37 deafferret :)
20:59 ank left #bioperl
21:00 ank joined #bioperl
21:06 vinnana joined #bioperl
21:11 flu left #bioperl
21:14 flu joined #bioperl
21:36 pyrimidine left #bioperl
21:57 carandraug left #bioperl
22:00 kyanardag left #bioperl
23:15 Yksi joined #bioperl
23:16 Yksi @dbolser hey! I'm guessing that you probably aren't at your computer atm, but just thought I'd let you know… i passed my MSc with Merit :D

| Channels | #bioperl index | Today | | Search | Google Search | Plain-Text | summary