gmane.comp.lang.perl.bio.general

Re: reorgansing tests by group

Chris Fields — 2008-12-04T03:39:13

You rock! (And consequently must be very tired). I'll try moving BioPerlTest into Bio::Root::Test (might be tricky)... chris On Dec 3, 2008, at 3:48 PM, Heikki Lehvaslaiho wrote:

Re: [Gmod-gbrowse] bioperl-live & annotator plugin

Lincoln Stein — 2008-12-02T15:49:45

Hi Sean, Did your plugin break when you upgraded to GBrowse 1.7, or when you upgraded bioperl? Knowing this will help me figure out where to look for the problem. It frankly looks like a database problem in the sense that the xyplot glyph is trying to render 100 features rather than a feature with 100 parts. Are you sure this step is correct? $feature_list->add_type($matrix=>{glyph => 'xyplot', It seems to me you intend to add the $feature, not the $matrix variable. Lincoln On Tue, Dec 2, 2008 at 10:32 AM, Sean O'Keeffe molgen.mpg.de>wrote:

Re: [Gmod-gbrowse] bioperl-live & annotator plugin

Lincoln Stein — 2008-12-02T15:51:06

I take it back. You should add the matrix in add_type(), assuming that $matrix is some sort of descriptive term. Lincoln On Tue, Dec 2, 2008 at 10:49 AM, Lincoln Stein gmail.com>wrote:

Re: [Gmod-gbrowse] Bioperl 1.6 and Bio::Graphics

Lincoln Stein — 2008-12-02T16:51:39

Hi Chris, I am supportive of splitting out Bio::Graphics, but not the Bio::DB::* modules before the 1.6 release. The only dependency issue is the module Bio::Graphics::FeatureBase, which is a lightweight hash-based Bio::SeqFeatureI module that is shared between Bio::Graphics::Feature and Bio::DB::SeqFeature. It should be renamed and kept in the main bioperl distribution. Will it confuse people if I rename this module Bio::SeqFeature::Lite? Lincoln On Wed, Nov 26, 2008 at 3:25 PM, Chris Fields illinois.edu> wrote:

Re: [Gmod-gbrowse] Bioperl 1.6 and Bio::Graphics

Chris Fields — 2008-12-02T17:07:43

Lincoln, Okay, we can do that. Would you want that to be under bio-graphics in subversion, or should we think ahead a bit in case other Gbrowse- dependent modules migrate there and call it something else? Bio::SeqFeature::Lite works for me. I can't see that causing too many problems as long as Gbrowse installation always picks the latest dev or CPAN release based on the version (though I don't know if there are any documentation fixes we need to take care of, maybe on the wiki). chris

Re: PullHSP bug

Chris Fields — 2008-12-03T02:40:32

It appears the problem boils down to a small inconsistency with the way blank IDs are handled between the two BLAST parsers. The regular blast parser explicitly sets this to '', while blast_pull leaves it undef. When one constructs a LocatableSeq w/o an ID it dies, but '' works, hence the problem. I don't think using '' is a real solution though (seems like something added to silence errors). The short and easy fix is either having get_nse() or blast_pull fall back to '', but both are technically wrong as there isn't a true ID. Ideas? Maybe a fallback 'unknown' ID (which seems to be used as a ). I don't think going back to the old behavior would be the correct thing to do. When adding seqs to a SimpleAlign, the NSE is used to index sequences and check for conflicts. The index was previously built from scratch (didn't use LocatableSeq::get_nse) but was recalled using get_nse, which is inconsistent. If get_nse changes slightly, such as optionally adding the version to rema

Re: on FootPrinter

Kevin Brown — 2008-12-02T17:35:06

Try installing a more recent version of BioPerl by following the directions at http://www.bioperl.org/wiki/Installing_Bioperl_for_Unix

About graphic rendering

Paolo Pavan — 2008-12-03T18:21:50

Hi everybody, I've a question about a Bio::Graphics rendering. I would like to display a blast HSP on a track but I would like to draw the "tails" of my mapping sequence that don't match the subject. So I instantiate these 3 object that represent all the three segments I want to display: my $feature = Bio::SeqFeature::Generic->new( -display_name =>$sequence_to_map->display_id, -start =>$hsp->sbjct->start, -end =>$hsp->sbjct->end, ); my $feature_pre = Bio::SeqFeature::Generic->new( -start =>($hsp->sbjct->start - $hsp->query->start), -end =>$hsp->sbjct->start-1 ); my $feature_post = Bio::SeqFeature::Generic->new( -start => $hsp->sbjct->end+1, -end

PullHSP bug

Chris Fields — 2008-12-02T23:03:21

Sendu, I found a silent bug while fixing some issues with SimpleAlign which are causing blast_pull tests to fail (HSP-related). It appears that the query sequence ID is not being caught, causing LocatableSeq::get_nse() to throw an exception within SimpleAlign when using PullHSPI::get_aln. It appears that the hit seq ID is caught properly, though (hence the oddness). All other SearchIO.t tests pass. Could you look into this? ... ok 165 ok 166 ok 167 ------------- EXCEPTION: Bio::Root::Exception ------------- MSG: Attribute id not set STACK: Error::throw STACK: Bio::Root::Root::throw /Users/cjfields/bioperl/bioperl-live/ blib/lib/Bio/Root/Root.pm:357 STACK: Bio::LocatableSeq::get_nse /Users/cjfields/bioperl/bioperl-live/ blib/lib/Bio/LocatableSeq.pm:301 STACK: Bio::SimpleAlign::add_seq /Users/cjfields/bioperl/bioperl-live/ blib/lib/Bio/SimpleAlign.pm:293 STACK: Bio::Search::HSP::PullHSPI::get_aln /Users/cjfields/bioperl/ bioperl-live/blib/lib/Bio/Search/HSP/PullHSPI.pm:471 STACK: t/SearchI

Re: PullHSP bug

Chris Fields — 2008-12-03T15:06:06

On Dec 3, 2008, at 5:38 AM, Sendu Bala wrote: ...fallback in some cases (Can tell I was getting tired). That solution requires get_nse to silently convert this; the exception is raised b/c the N/S-E is used for indexing (so lack of any one of those may result in indexing issues, such as replacing sequences with same ID but no S-E, etc.). We don't want the exception turned off unless it is explicitly known the sequence instance lacks any or all of the N/S-E, such as in this case. Setting an explicit LocatableSeq flag for such instances (force_nse), then having get_nse() DTRT internally would be easy to implement. Odd, list mail hasn't been coming through (bioperl-gut either). I'll post something to support. chris

Fw: Help with installation

Assiatu Barrie — 2008-12-02T17:40:31

Assiatu Barrie Dept. of Biology Univ of TX, Arlington Arlington, TX 76019 abarrie< at >uta.edu assiatu_barrie< at >yahoo.com www.uta.edu --- On Tue, 12/2/08, Assiatu Barrie yahoo.com> wrote: From: Assiatu Barrie yahoo.com> Subject: Help with installation To: bioperl-l< at >biolperl.org Date: Tuesday, December 2, 2008, 11:36 AM Hello My name is Assiatu Barrie, a grad student from UTA. I am trying to download bioperl on my mac os laptop but unsuccessful, please help. I keep getting an error message regarding Make not been installed. I tried installing Build.pl, but not successful. I am very very new to perl so I will really appreciate the help. Thank you Assiatu Barrie Dept. of Biology Univ of TX, Arlington Arlington, TX 76019 abarrie< at >uta.edu assiatu_barrie< at >yahoo.com www.uta.edu

Re: PullHSP bug

Sendu Bala — 2008-12-03T11:38:30

I'd say for now at least the fix would be to alter get_nse() to fall back to ''. You don't want the work-around code duplicated in multiple places. Indeed, the normal blast parser should be altered to return undef as well. Then if a more desirable solution is thought of in the future, there's just one place to implement it: in get_nse().

Re: PullHSP bug

Sendu Bala — 2008-12-03T00:25:21

I don't see this as a bug of PullHSPI. The blastn report being parsed in that failing test has no query sequence id, so obviously PullHSPI can't give an ID to the query sequence it creates. Why does SimpleAlign now fall over when an added sequence has no id? I'd suggest the old behaviour be retained...

global verbosity

Mark A. Jensen — 2008-12-03T15:23:32

I was looking for a way to set verbosity globally (but not for debugging), so that all subsequently created objects would be absolutely quiet. I found $Bio::Root::Root:VERBOSITY, but saw that it didn't matter to downstream objects when it was changed. Digging in to the issue, found that global verbosity can't really be set, though one can access the value of $VERBOSITY with Bio::Root::Root::verbose(). http://bugzilla.open-bio.org/show_bug.cgi?id=2696 has a patch + .t to Root.pm that I think solves the issue, allowing getting and setting of global verbosity, so that the global value now propagates to new objects. There is a stickier issue (example using Bio::Seq and Bio::Seq->primary_seq() in the .t file), in which setting an object's verbosity post-construction does not propagate to sub-objects contained in the object, so that one might not get what one expects. cheers MAJ

Re: Fw: Help with installation

Mark A. Jensen — 2008-12-03T16:32:12

Assiatu- You may find the fink installer helpful- see http://www.bioperl.org/wiki/Getting_BioPerl#MacOS_X_using_fink cheers MAJ

Re: reorgansing tests by group

Heikki Lehvaslaiho — 2008-12-03T18:23:28

2008/12/2 Chris Fields illinois.edu>: Chris, I'll try to do the rests very soon now. Do not touch the t directory until I've committed by reorganisations! Moving to core would be a good thing. Do it. -Heikki

Re: reorgansing tests by group

Heikki Lehvaslaiho — 2008-12-03T21:48:06

2008/12/3 Chris Fields illinois.edu>: I jsut had long flight that gave me time to think this through! Now they are done! The ones on the top level are deprecated, unfinished or can be left there (Perl.t & PodSyntax.t). I moved some tests from t/Tools into t/SeqTools. All of them give additional functionality to Sequence objects and are not parsers like tests in t/Tools directory. The exceptions in current t/Tools are Analysis subdirectory tests. I would have pulled them and t/Tools/Run/RemoteBlast.t into t/WebAnalysis.t I committed all in one go. I suspect that the mail generated is too long (like in your case) and does not get sent. -Heikki

Re: reorgansing tests by group

Chris Fields — 2008-12-02T13:37:44

Wow, I go to sleep and the tests are almost reorganized. Thanks Heikki! The only (very small) caveat I find to using recursive tests is any additional non-test files we add, including data for tests in the / data directory, must not end with *.t, otherwise the build file will try to run it and bomb. There was one file I changed to .tt which is now passed over. If needed we can move over to an autogenerated file which explicitly lists the tests to run (this would likely be at a later point). On a related point, what does everyone think of moving BioPerlTest to core so the other distributions can use it? I think it is worth doing eventually, but it could possibly wait until 1.6.x or 1.7. I have tested it out and it works but needs some dependency issues ironed out. chris On Dec 2, 2008, at 12:05 AM, Heikki Lehvaslaiho wrote:

Re: reorgansing tests by group

Chris Fields — 2008-12-03T20:03:27

On Dec 3, 2008, at 12:23 PM, Heikki Lehvaslaiho wrote: ...also, just to note, the mail lists appear to be acting up (support has been notified). My commits last night are here: http://lists.open-bio.org/pipermail/bioperl-guts-l/2008-December/date.html chris

Re: reorgansing tests by group

Chris Fields — 2008-12-03T19:53:31

On Dec 3, 2008, at 12:23 PM, Heikki Lehvaslaiho wrote: Okay. I moved Tools already but I'll leave everything else. That'll be next then. -c

Re: reorgansing tests by group

Heikki Lehvaslaiho — 2008-12-02T06:05:15

Yes, forgot that! The directory needs to be added to the repository: 1. mkdir group 2. svn add group 3. svn move test1.t. group 4. svn move test2.t. group 5. ... -Heikki 2008/12/2 Hilmar Lapp gmx.net>: