Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-07-22

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

All times shown according to UTC.

Time Nick Message
00:01 carandraug joined #bioperl
03:15 dukeleto joined #bioperl
03:21 dukeleto joined #bioperl
06:17 bag joined #bioperl
13:10 ank joined #bioperl
13:24 brandi joined #bioperl
13:24 brandi left #bioperl
14:13 ank joined #bioperl
15:17 dmb_ joined #bioperl
15:18 Guest36638 deafferret: AWS Public Data
15:19 faceface Cloud copies of Ensembl, GenBank, 1000 Genomes and many many more.
15:56 pyrimidine joined #bioperl
16:20 JunY Hi, does anyone know how to control the format of POD file...
16:21 JunY Sometimes, I just feel, after I use perldoc, the POD instruction is too ugly... really bad formated...
16:24 pyrimidine JunY: what do you mean?  The POD markup?
16:25 JunY I just want the perl .pm file to be well formated
16:25 pyrimidine I have taken up the habit of pushing all POD after the __END__ of the module
16:26 pyrimidine but that's me, and opinions vary on that style
16:26 JunY Ok, I will check that later.
16:26 pyrimidine It does tend to keep all documentation together in one place, and for me that's easier to format
16:27 JunY A good news. Retrieving alignment file from Pfam is partially finished.
16:27 JunY I have written a get_Aln_by_acc method in Bio::DB::Align::Pfam
16:27 CIA-95 bioperl-live: Duke Leto topic/eutils_carryover * r19d5b75 / t/Ontology/IO/interpro.t : Refactor some tests in t/Ontology/IO/interpro.t - http://bit.ly/cmMBmw
16:27 CIA-95 bioperl-live: Duke Leto topic/eutils_carryover * r7314afb / (3 files in 3 dirs): Add Conserved_site as an InterPro root term and update relevant tests - http://bit.ly/9MEKJQ
16:27 CIA-95 bioperl-live: Chris Fields topic/leto_interpro_get_dbxref * r5054528 / Bio/OntologyIO/Handlers/InterProHandler.pm : squash warnings popping up with uninit data; marked as a TODO (is this supposed to be an error?) - http://bit.ly/d7Ft6J
16:28 JunY Now, you can just do: my $dbobj=Bio::DB::Align->new(-db=>"pfam"); my $aln=get_Aln_by_acc("PF00001");
16:29 JunY Then, $aln will be the Bio::SimpleAlign object you want
16:29 pyrimidine should that be $dbobj->get_Aln_by_acc?
16:29 JunY yep, you are right
16:29 JunY my $aln=$dbobj->get_Aln_by_acc()
16:29 pyrimidine looks good.
16:30 JunY I will move on to Uniprot tomorrow
16:30 pyrimidine cool!
16:31 JunY Hopefully, this new package will support pfam, uniprot, pdb and embl databases... we will see
16:31 JunY And maybe alignment file from a few other databases... if they provide RESTful service
16:33 JunY The only problem is that, in Pfam, this method is only limited by searching using accession. It cannot do the id search, e.g. get_Aln_by_id, at the moment. But I will check that later. Probably, I need to write a full package, e.g. Bio::DB::Pfam to interpret the id to accession tranformation.
16:38 pyrimidine wouldn't worry too much about that (ID-based search).  Can always throw_not_implemented
16:39 pyrimidine dukeleto: commit to your branch was just to squelch some warnings popping up.  Don't know if those should be errors or not...
16:40 dukeleto pyrimidine: i will look at it
16:41 pyrimidine afk, lunch&
16:43 dukeleto pyrimidine: it could be caused by the sample interpro file referencing interpro terms that are not included in the same file
16:43 JunY ok, see u guys tomorrow
16:44 JunY Hasta la vista
18:17 pyrimidine dukeleto: looks like Bio::Ontology::InterProTerm falls back to using the parent class get_dbxrefs() (Bio::Ontology::Term)
18:18 pyrimidine that appears to be pointing to the wrong place for stored dbxrefs for InterProTerm
18:30 dukeleto pyrimidine: yes, i noticed it was inheriting the method, but didn't get further than that
18:31 dukeleto pyrimidine: do InterProTerm's need a new get_dbxref method, or can we steal it from somewhere else?
18:36 pyrimidine probably need to override get_dbxrefs, yes
18:36 pyrimidine unfortunately
18:37 pyrimidine dbxrefs seem to be stored in a Bio::Annotation::Collection
18:38 pyrimidine oops, take that back
18:39 pyrimidine they are stored in various attributes
18:40 pyrimidine so you would have to either decide which of those to return (members, examples, external_documents, class_list)
18:40 pyrimidine or get_dbxrefs would return them all
18:41 pyrimidine unless I'm misreading this...
18:41 pyrimidine (entirely possible :)
18:44 pyrimidine Bio::OntologyIO::Handlers::InterProHandler, lines 663, 682, 694, 705
18:44 pyrimidine that's where they're being added to the terms
18:46 pyrimidine reading get_dbxrefs(), it seems you can specify the context
18:46 pyrimidine so, you could do something like get_dbxrefs() to get everything
18:48 pyrimidine but specify get_dbxrefs('members') to just get those
18:53 dukeleto pyrimidine: interesting
18:55 pyrimidine seems like there an inconsistency in the API there
18:59 dukeleto pyrimidine: i am trying it and get_dbxrefs('members') still returns undef
19:04 pyrimidine I think we would have to override get_dbxrefs for InterProTerm to grab the right ones
19:10 dukeleto pyrimidine: yeah, i see what it is doing now
19:10 dukeleto pyrimidine: i will write a bunch of tests for each kind of dbxref, then i will start hacking on that
19:11 pyrimidine sounds
19:11 pyrimidine *sounds good
19:11 pyrimidine for some reason, there are separate getter/setters for each type
19:16 dukeleto pyrimidine: do we want to keep it that way?
19:16 pyrimidine probably (backwards compat), but add deprecation warnings to them
19:17 pyrimidine seems to make more sense to add them consistently within InterProHandler
19:20 pyrimidine end_element there is looking a bit, erm, messy, but it works
19:20 * pyrimidine just loves very long if-elsif blocks
19:22 dukeleto pyrimidine: yeah, i hear ya. bioperl6 must be fun to hack on, since you don't have to worry about backcompat and you can do everything The Right Way
19:23 pyrimidine it is fun
19:23 pyrimidine now if I only had tuits for it :(
19:23 pyrimidine biome is the same way (bioperl in Moose)
19:38 * dukeleto is amazed that pyrimidine has time to hack on bioperl, biome and bioperl6
19:39 pyrimidine not as much anymore
19:39 pyrimidine but we have someone who is contributing to bioperl6 (takadonet, on #perl6), and I've been putting a bit more into biome
19:40 pyrimidine would like to get bioperl to a point where releases are easy, so thinking something like Dist::Zilla or ShipIt is in our future
19:43 deafferret faceface: do you have a website showing how to mount the EC2 AWS?
19:52 dukeleto pyrimidine: i have heard very good things about Dist::Zilla and I am planning to convert my CPAN modules over to it
19:54 dukeleto pyrimidine: should 'deleted_entries' be a dbxref context as well?
19:54 * pyrimidine looking
19:55 pyrimidine seems like those are just Ontology IDs that are no longer used
19:57 pyrimidine looks like the handler marks particular those InterProTerms as obsolete
19:57 dukeleto pyrimidine: yeah, I am just trying to make sure what the complete list of valid contexts should be. Right now I am thinking : class_list, external_doc_list, example_list, sec_list, member_list, pub_list
19:57 dukeleto pyrimidine: does that look correct? should 'cite' be a context as well?
19:58 pyrimidine looks like 'cite' is a Bio::Annotation::Reference
20:02 dukeleto pyrimidine: so that is a 'no', correct ? :)
20:02 pyrimidine that is a 'no'
20:02 * dukeleto is still getting used to all of the bioperl terminology
20:02 pyrimidine it's a bit, erm, messy
20:03 * pyrimidine is wondering why all the Bio::Annotation stuff isn't handled with an Bio::Annotation::Collection
20:03 pyrimidine at least that would be consistent with everything else in bioperl
20:05 CIA-95 bioperl-live: Jonathan "Duke" Leto topic/leto_interpro_get_dbxref * r09f5f0e / t/Ontology/IO/interpro.t : Add a test for each kind of InterPro dbxref context - http://bit.ly/dafER8
20:12 buelsbot joined #bioperl
20:46 pyrimidine dukeleto: I made some changes to the handler and InterProTerm to simplify things, but the '_dblinks' hash key is getting clobbered somewhere
20:47 pyrimidine it's odd
20:47 pyrimidine want me to commit these?  Has some debugging...
20:52 bag_ joined #bioperl
20:57 CIA-95 bioperl-live: Jonathan "Duke" Leto topic/cjfields_interpro_get_dbxref * r09f5f0e / t/Ontology/IO/interpro.t : Add a test for each kind of InterPro dbxref context - http://bit.ly/dafER8
20:57 pyrimidine even easier, will make a branch from your branch with the changes
21:00 CIA-95 bioperl-live: Chris Fields topic/cjfields_interpro_get_dbxref * r22c4108 / Bio/Ontology/InterProTerm.pm : reimplement various methods in terms of get_dbxrefs, for consistency - http://bit.ly/b60pIc
21:00 CIA-95 bioperl-live: Chris Fields topic/cjfields_interpro_get_dbxref * r5fb02a9 / Bio/OntologyIO/Handlers/InterProHandler.pm : use add_dbxrefs with contexts instead of the various methods (could revert back, but this is more direct) - http://bit.ly/9MNFQv
21:01 deafferret inter...PRO!!!
21:01 * deafferret swims around in MATLAB licensing server for the first time
21:02 rbuels eeeww?
21:02 deafferret CIA-95: rbuels is also just jealous
21:03 deafferret ALL YOUR MATHS ARE BELONG TO ME
21:05 pyrimidine I live in the land of Wolfram, we can't say MATLAB w/o fear of the Mathematica cops coming after us
21:06 deafferret wolfram?
21:06 pyrimidine http://www.wolframalpha.com/
21:06 deafferret ma***b, the vendor whose name shall not be spoken
21:07 pyrimidine http://www.wolfram.com/
21:07 deafferret tee hee   http://www.wolframalpha.com​/input/?i=matlab+licensing
21:07 pyrimidine "Development of this topic is under investigation"
21:07 pyrimidine See, I'm telling you, they're watching you
21:08 * deafferret only fears websites with flashlights under their faces
21:09 dukeleto pyrimidine: the things you are chaning in your interpro topic branch don't have any tests, do you want me to add some to your branch? i was going to add some, but was thinking if my topic or master was better
21:10 pyrimidine http://www.youtube.com/watch?v=Bmk9CjEha8A
21:10 pyrimidine dukeleto: your branch or master
21:10 pyrimidine I can merge/rebase from there
21:11 dukeleto pyrimidine: ok
21:12 * pyrimidine now can't get Conan's 'In the year 3000' out of his head...
21:14 pyrimidine dukeleto: in general, all these changes do is make the InterProTerm-specific methods delegate to get/add/remove_dbxrefs as needed
21:15 pyrimidine also noticed that reference code is essentially the same as the parent class, so commented those out.  Seems to pass tests, except the dbxref ones :P
21:16 dukeleto pyrimidine: nice
21:16 dukeleto pyrimidine: we are going to need a bunch more test data for the dbxref tests
21:16 pyrimidine add away
21:17 pyrimidine I'll plow ahead on working out where dbxrefs are tanking
21:21 * dukeleto just noticed that root terms don't have dbxrefs, but leaf terms do
21:21 dukeleto i still think there is a bug, but i may not need to add much more test data
21:26 pyrimidine yep, that's what's happening
21:31 pyrimidine that's an odd design, actually
21:31 pyrimidine so, the Term has a link to the ontology
21:32 pyrimidine but it appears the Term in this case isn't actually part of the ontology
21:34 pyrimidine http://gist.github.com/486623
21:35 pyrimidine Dump of example InterProTerm
21:35 CIA-95 bioperl-live: Jonathan "Duke" Leto topic/leto_interpro_get_dbxref * r3a817b6 / t/Ontology/IO/interpro.t : Add some tests for get_members and protein_count and refactor tests for get_dbxrefs - http://bit.ly/9V2rnF
21:35 pyrimidine all of the data is stuck in the ontology
21:36 dukeleto yeah, everything carries it's ontology around. it makes looking at Data::Dumper output a mess
21:36 pyrimidine the dbxrefs in question are in the ontology, under _term_store
21:36 pyrimidine which explains why we don't see anything, at least
21:37 pyrimidine I don't think the other ontologies are implemented in this way, but I'm not sure
21:42 CIA-95 bioperl-live: Jonathan "Duke" Leto topic/leto_interpro_get_dbxref * r6c2c954 / t/Ontology/IO/interpro.t : Add tests for class_list, get_examples and get_external_documents - http://bit.ly/cvO4rh
21:42 pyrimidine example of GOTerm: http://gist.github.com/486636
21:42 pyrimidine note the dblinks are present
21:43 pyrimidine so something is amiss
21:45 * dukeleto notes that something is rotten in the state of bioperl
21:47 pyrimidine this is the dark side of 'the cathedral and the bazaar', where we're the bazaar
21:51 pyrimidine lots of different programmers, lots of different styles
21:52 rbuels (and very little review for a long time)
21:53 pyrimidine yep
21:55 pyrimidine so, my guess is the handler is always pulling the term from the ontology for storing things, and the root term is simply passed over
21:57 CIA-95 bioperl-live: Jonathan "Duke" Leto topic/leto_interpro_get_dbxref * rc495ed1 / t/Ontology/IO/interpro.t : Add a test for get_references and to_string - http://bit.ly/bKWcZL
22:00 pyrimidine I definitely think this is a good area for documentation.  There are very little (any?) docs/HOWTOs for Bio::Ontology, even though they're extensively used
22:00 dukeleto yes, there were basically no docs for any ontology stuff, i just read the source
22:00 pyrimidine which is why we're seeing variations on the implementation
22:01 pyrimidine (by variations, I mean bugs)
22:02 * deafferret wonders if he's violating any MATLAB privacy / licensing agreements... http://biodoc.ist.unomaha.edu/wiki/MATLAB
22:02 deafferret / confidentiality
22:02 * deafferret frames dukeleto
22:04 * pyrimidine finds it ironic that UIUC has a free MATLAB license
22:05 pyrimidine or maybe that's strategic
22:07 deafferret py[tab] needs a cool photo of him fencing or something    http://illinois.edu/ds/detail?userId=cjfiel​ds&search_type=staff&search=*%20fie​lds&from_result_list=true&skinId=0#
22:09 deafferret boo!  The IGB webcam is currently unavailable due to building maintenance.
22:11 pyrimidine http://www.igb.uiuc.edu/people/IGBfellows.html
22:11 * dukeleto wonders what the framing was about
22:12 deafferret dukeleto: MathWorks' lawyers should be contacting you later this week
22:12 deafferret sorry about that
22:13 deafferret pyrimidine: hey! there you are!
22:13 dukeleto deafferret: lulz
22:15 pyrimidine :)
22:15 deafferret i'm surprised they don't close the campus when the webcam is down. what's the point if the Internet can't watch?
22:16 * deafferret definately doesn't want to miss the next webcam flash mob
22:18 pyrimidine They turned the camera off a few months back, I think it was hacked
22:19 deafferret hacked how? it was suddenly shooting out lasers killing people?
22:19 deafferret tsk tsk
22:20 pyrimidine someone set the admin password to something easily broken, my guess
22:20 deafferret and... what? looked at the pictures? who cares?
22:20 deafferret turned it off?
22:21 * deafferret puts a big sign on it "Feel free to hack this, dorks"
22:21 pyrimidine you can remote control it via web browser
22:21 pyrimidine so my guess is someone thought it was a challenge
22:21 pyrimidine you can always have this: http://bit.ly/aBqrDO
22:22 pyrimidine too bad it isn't live
22:22 deafferret i.... don't understand. who gives a crap if someone hacks the camera? let them hack it
22:23 pyrimidine I don't care, personally.  It's the admins here who do.  Stick up asses and all.
22:25 rbuels it's the principle of the thing.
22:26 deafferret rbuels: do tell: what principal would that be then?
22:27 deafferret "i have to look like I'm in control of meaningless shit" ?
22:27 CIA-95 bioperl-live: Chris Fields topic/cjfields_interpro_get_dbxref * r846480a / Bio/Ontology/Term.pm : a bit o' debugging; alias add_dbxrefs to add_dbxref - http://bit.ly/9h4Tog
22:27 CIA-95 bioperl-live: Chris Fields topic/cjfields_interpro_get_dbxref * re867641 / (2 files in 2 dirs): pinpointing the problem, perhaps, paint me perplexed - http://bit.ly/aFyDgR
22:27 pyrimidine deafferret: pretty much
22:27 rbuels principal Belding, from saved by the bell.
22:28 * rbuels feeds a gnome
22:29 pyrimidine we had some local high school kids hack the automated calling system for school cancellations
22:29 deafferret doh!
22:29 pyrimidine pretty funny when it happened
22:29 rbuels oh ho ho ho
22:29 deafferret schools out for summer ... school's out FOREVER
22:30 pyrimidine it went a bit further than that
22:30 pyrimidine more along the lines of: "these teachers suck, X Y Z"
22:31 pyrimidine "yours truly, the superintendent"
22:31 * deafferret tussles the scamps hair
22:32 rbuels gee whiz, what'll we ever do with these kids
22:32 deafferret i have a pretty high tolerance for that stuff, short of theft, damage, or threats of violence
22:34 deafferret hmm... think i'll hit the cafeteria before it closes
22:34 pyrimidine http://www.reddit.com/r/programming/comments/crybo​/rakudo_stars_really_really_release_ready_porting/
22:39 * pyrimidine commuting
22:40 pyrimidine o7
22:40 deafferret toodles
23:41 dukeleto exit

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