Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-07-30

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

All times shown according to UTC.

Time Nick Message
00:23 dnewkirk joined #bioperl
01:53 dnewkirk joined #bioperl
02:43 dnewkirk left #bioperl
04:27 CIA-94 bioperl-live: Chris Fields master * r4ee526b / (Bio/Tools/GuessSeqFormat.pm Changes): [bug 3122] catch seek calls on the unseekable (thanks to Stefan and Roy) - http://bit.ly/awARP5
04:47 CIA-94 bioperl-live: Chris Fields master * rb31f535 / (Bio/DB/Fasta.pm Changes t/LocalDB/DBFasta.t t/data/dbfa/1.fa): [bug 3126] catch description (patch courtesy Toshihiko Akiba); test added - http://bit.ly/98ScS9
05:02 CIA-94 bioperl-live: Chris Fields topic/bug3039 * r15aad7e / (Bio/TreeIO/newick.pm t/Tree/TreeIO/newick.t): [bug 3039] added Dave's changes to a branch; these are causing test failures outside of t/newick.t, unfortunately, so needs investigating - http://bit.ly/90gCSI
05:04 bag_ joined #bioperl
07:38 vinnana joined #bioperl
10:53 brandi joined #bioperl
10:53 brandi left #bioperl
12:38 kai haha
12:39 kai on OpenSuse Factory, my BioPerl rpm builds
13:09 jhamilton joined #bioperl
13:17 dbolser hello.
13:17 dbolser after the merge should I ...
13:17 dbolser I think I should branch my current codebase, and rebase on bioperl-live?
13:18 * dbolser wonders how he ended up here again ;-)
13:20 dbolser git checkout -b my_new_branch
13:20 dbolser Switched to a new branch "my_new_branch"
13:20 dbolser git checkout master
13:21 dbolser Switched to branch "master"
13:21 dbolser Your branch is ahead of the tracked remote branch 'origin/master' by 20 commits.
13:21 dbolser so if I rebase, I'll lose those 20, but thats fine, because they are tucked away under 'my_new_branch'
13:21 dbolser ?
13:22 dbolser or how should I reset my codebase to origin/master?
13:22 dbolser (bioperl-live)
13:31 dbolser git reset --hard origin/master :D
14:11 dbolser pyrimidine: just tried to tests latest codebase
14:11 dbolser found an un-reported dep for GD::Group
14:11 dbolser in graphics
14:11 dbolser had to install gd-devel and cpan GD::Group
14:31 dbolser config for BES?
14:32 dbolser gbrowse
14:32 * dbolser goes to gmod
14:36 Jun Hi, did anyone use Bio::AlignIO::Pfam recently? It may have a bug...
14:37 dbolser Jun: not me
14:37 dbolser can you write a test case?
14:44 Jun ok...
14:45 Jun http://rfam.sanger.ac.uk/family/alignment/dow​nload/format?acc=RF00360&alnType=seed&amp​;nseLabels=1&format=pfam&download=0
14:45 Jun Here is an pfam alignment file from rfam. If you use Bio::AlignIO::pfam, it will give a warning and stop to run ...
14:53 Jun Or, maybe a quick question. Does "pfam alignment format" eq "stockholm alignment format"?
15:01 dbolser I think they are related
15:01 dbolser but not 100% sure if that is true
15:01 dbolser ask in #bioinformatics ?
15:04 Jun ok, i will try. Thx
15:04 Jun join #bioinformatics
15:13 deafferret .oo.
15:26 dbolser yo deafferret
15:26 dbolser do you consult on configuring gb, and if so, how much is 4 hours of your time ;-)
15:27 pyrimidine Jun: pfam != stockholm
15:27 pyrimidine I think pfam is simpler
15:29 pyrimidine dbolser: rebase should replay your changes over those on master, you shouldn't lose them
15:31 Jun Yep, I tried to run Bio::AlignIO::pfam on a pfam alignment file from rfam. It failed.
15:31 Jun I think Bio::AlignIO::pfam may need to be rewritten... or just redirected to Bio::AlignIO::Stockholm
15:36 pyrimidine might be a bad idea, looking at it
15:37 pyrimidine there is a subtle difference btwn the two, which is terrible, actually (b/c the tag at the top says it's stockholm, not pfam)
15:37 Jun check this:http://rfam.sanger.ac.uk/family/alignment/dow​nload/format?acc=RF00360&alnType=seed&amp​;nseLabels=0&format=pfam&download=0
15:38 Jun This is a pfam format from Rfam. But the module failed in reading it ...
15:39 pyrimidine http://rfam.sanger.ac.uk/family/alignment/dow​nload/format?acc=RF00515&alnType=seed&amp​;nseLabels=0&format=pfam&download=0
15:39 paul__ joined #bioperl
15:39 pyrimidine http://rfam.sanger.ac.uk/family/alignment/down​load/format?acc=RF00515&alnType=seed&n​seLabels=0&format=stockholm&download=0
15:39 paul__ hi
15:39 paul__ can somebody help me with a write_seq question?
15:40 pyrimidine paul__: paste to the pastebin (read the IRC topic)
15:40 paul__ how do I output a fasta sequence so that the sequence is printed on one line? In other words, I don't want the sequence to wrap
15:41 paul__ where is pastebin?
15:41 paul__ sorry i'm quite new
15:41 pyrimidine http://pastebin.com/
15:41 pyrimidine use /topic to find it
15:41 pyrimidine paste the link here
15:43 paul__ http://pastebin.com/nLcNygrW
15:43 Jun Hi, pyrimidine. Bio::AlignIO::pfam actually work on the old file:http://github.com/bioperl/bioperl-l​ive/blob/master/t/data/testaln.pfam
15:43 pyrimidine paul__: I meant, if you have example code, paste it there.
15:43 Jun ok, I think I know paul_'s answer
15:44 pyrimidine pfam, IIRC, is just stockholm w/o interleaving
15:44 Jun paul__, maybe you can try $aln->write_aln(-width=$aln->length)
15:44 Jun paul__, maybe you can try $aln->write_aln(-width=>$aln->length)
15:44 paul__ that seems too complicated
15:45 pyrimidine paul__: why is that complicated?
15:45 pyrimidine he gave you an answer.
15:45 pyrimidine it should work
15:45 paul__ i'm not working with alignments
15:45 pyrimidine ah
15:45 paul__ i have a single fasta sequece
15:45 paul__ that is on multiple lines
15:46 paul__ i want to write a script to put all the sequence info on one line
15:46 paul__ i was hoping that Bio::SeqIO write_seq would output all sequence on one line
15:46 pyrimidine I think there is an analogous way of doing the above for SeqIO
15:46 paul__ but it wraps the sequence
15:46 pyrimidine yes, b/c that's the general standard for FASTA
15:46 pyrimidine so it's the default
15:47 paul__ so there is no option to specify fasta width?
15:47 paul__ like width = 0 could be one line output
15:48 paul__ how hard would it be to add that type of functionality?
15:48 pyrimidine very easy
15:48 paul__ what do I need to learn about BioPerl in order to do that? What modules should I look at?
15:48 paul__ where is the code for write_seq?
15:49 pyrimidine Bio::SeqIO::fasta
15:50 pyrimidine couple of lines which would likely need to be changed
15:50 pyrimidine $str =~ s/(.{1,$width})/$1\n/g;
15:50 Jun Ok, it seems you can defined the width of the output by $seq->width
15:50 pyrimidine yep
15:50 pyrimidine $str =~ s/(.{1,$width})/$1\n/g if $width;
15:50 pyrimidine should work
15:50 Jun So paul__, you can try $seq->width($seq->length); $seq->write_seq();
15:51 pyrimidine but, it needs tests!
15:51 paul__ i'll try Jun's suggestion...thank you
15:51 paul__ Jun, where did you find the "width" method? What documentation?
15:51 Jun yep, try it. If you have any problem, come back
15:51 pyrimidine paul__: it's in Bio::SeqIO::fasta
15:51 Jun Bio::SeqIO::fasta
15:52 paul__ ah, I was only looking at Bio::SeqIO
15:52 Jun It seems, Bio::SeqIO has a similiar design as Bio::AlignIO.
15:52 paul__ i forgot there are specific submodules of SeqIO
15:52 Jun The implementations are all in Bio::SeqIO/ folder
15:52 pyrimidine Jun: other way around, actually
15:52 paul__ Thank you everybody
15:53 Jun np
15:53 pyrimidine paul__: if you patch this let us know, we can make the change to the core dist
15:54 paul__ ok but I think the "width" method is all I need....i'll try and be right back
15:55 Jun pyrimidine: it seems the pfam format has been changed... The current Bio::AlignIO::pfam was written 10 years ago. I wouldnot be surprised if it doesnot work
16:01 CIA-94 bioperl-live: Robert Buels master * rb0b52cc / (2 files): dir in maint for working on dynamiting the monolith - http://bit.ly/dckbDJ
16:02 * rbuels stacks coils of primacord
16:02 * rbuels whistles merrily
16:21 paul__ the "width" function was exactly what I needed...ty again
16:26 pyrimidine paul__: nice to hear
16:32 pyrimidine Jun: I think Pfam is breaking on the extra alignment annotation (not expecting it).  Problem is, redirecting to Stockholm will likely break older Pfam parsing
16:32 pyrimidine b/c stockholm actually expects the annotation and the "#STOCKHOLM" part
16:51 ank joined #bioperl
16:52 kblin rbuels: but you're not starting pre 1.6.2, are you?
16:53 * kblin stacks some sandbags, just in case
16:53 rbuels kblin: i'm taking the approach of doing most of the work of the split by constructing metadata about where things will go.  then i will write some code to do the split and probably automate running tests
16:54 pyrimidine rbuels: that makes the most sense to me
16:54 kblin sounds reasonable, yeah
16:54 rbuels kblin: this will avoid any merging problems, the hell of moving things around when mistakes are found, etc etc.
16:54 kblin ok
16:54 rbuels more like, try the split, see what breaks, tweak the splitspec, try it again, etc
16:54 kblin bbiaf, need to go shop for some foodstuffs
16:54 pyrimidine still working from Bio::Root up, correct?
16:55 rbuels pyrimidine: well actually it's going to be all at the same time.
16:55 rbuels pyrimidine: i think.
16:56 pyrimidine ok
16:56 rbuels pyrimidine: well, realistically, there will probably be some manual tweaking, and that will proceed from the bottom up
16:57 rbuels pyrimidine: the .csv file in there right now is a listing of every file in the bioperl-live, and when dist it will eventually go into
16:57 rbuels s/when/which/
16:58 rbuels pyrimidine: feel free to tweak it if you wind up looking at it
17:01 pyrimidine so, basically, we need to work out dists
17:01 CIA-94 bioperl-live: Nathan Liles master * rc1abbd1 / (Bio/Tools/GFF.pm Build.PL): Force ID field for all gb files converted to gff3 - http://bit.ly/9PgfW4
17:04 * deafferret basically needs to work out more
17:09 rbuels pyrimidine: yeah.  and the dist assignments will need to be tweaked as weird dependency things come to light.
17:09 rbuels pyrimidine: i chose csv cause it's easy to parse and easy to diff
17:10 pyrimidine makes sense to me
17:10 pyrimidine which reminds me
17:10 rbuels pyrimidine: and a spreadsheet actually makes a good editor for this
17:10 pyrimidine mark's bioperl-mode tarballs need manual merging over to 1.6
17:13 * rbuels does not get it
17:13 pyrimidine the tar files?
17:14 pyrimidine they were in conflict, but binary merge is a PITA
17:14 pyrimidine so, just simple 'git checkout master foo.txt' while in 1.6
17:14 rbuels oh, makes sense
17:14 rbuels seems to me those should not be tars.
17:15 pyrimidine agreed
17:15 pyrimidine not sure how big they are untarred
17:15 pyrimidine de-tarred?
17:15 rbuels same size! it's not compressed
17:22 * deafferret plays suspenseful music
17:24 CIA-94 bioperl-live: Chris Fields branch-1-6 * r1766d4e / (9 files in 6 dirs): Merge branch 'master' into branch-1-6 (+5 more commits...) - http://bit.ly/blDwOI
17:30 * deafferret creshendos
17:30 rbuels deafferret: crescendo!  get a spelchecking irc client!
17:30 rbuels gosh!
17:30 deafferret spelchecking is enabled
17:31 * deafferret clicks the spelunking button, disappears into a hole
17:32 deafferret besides, i've been clicking 'add to dictionary' for years. don't blame me that YOUR spellchecker dictionary is out of date   :p
17:32 * deafferret chuckles, a gnome falls out
17:32 * deafferret lools all over the place
17:35 rbuels wow, you're a just a big bundle of inside jokes today
17:35 * rbuels feeds the phant
17:35 deafferret these are all #bioperl classics
17:35 deafferret that we have all enjoyed for eons
17:36 rbuels they were good enough for my father, and his father before him
17:37 pyrimidine left #bioperl
17:38 deafferret "This record has been skillfully crafted by British comedians using ancient, well worn, classical, hand tooled jokes. It has been specially designed to sit at the back of your record collection amongst the old Frank Sinatra albums to be brought out and split up when you get divorced. Any complaints about the humorous quality of this album should be addressed to British Airways, Ingraham's Drive, Greenwich."
17:42 pyrimidine joined #bioperl
17:47 rbuels deafferret: haha, where is that from
17:51 deafferret http://en.wikipedia.org/wiki/The_Final_Rip_Off
17:52 spekki0 joined #bioperl
18:02 pyrimidine deafferret: I have that somewhere, buried in my CD collection
18:03 deafferret I bought it again a couple years ago. too many memories from college, then lost it somewhere  :)
18:03 deafferret (iTunes)
18:03 deafferret pyrimidine: make sure you keep it in the divorce
18:04 pyrimidine hah.  Don't think that's happening, unless my wife forgot to tell me something...
18:04 spekki0 I'm having some issues with using Bio::DB::EUtilities efetch, whenever i post a gi number that has been removed such as 157749793 i get the following error when trying to download the record. http://codepad.org/IoVhxTeB, i have the line $gpeptfactory->get_Response in a eval block but it still doesnt seem to be handling it very well. im just wondering does something still get downloaded when the record has been removed from ncbi's site or do
18:04 spekki0 es it just fail?
18:04 pyrimidine :P
18:05 deafferret pyrimidine: oh, you didn't know? sorry, man
18:05 deafferret >;)
18:05 pyrimidine happens to the best of us
18:05 deafferret divorce?
18:05 pyrimidine spekki0: that's on NCBI's end
18:05 pyrimidine Internal Server Error
18:05 deafferret thank god i suck so 'the best of us' does not apply
18:06 deafferret Bio::NCBI::Fixer
18:06 spekki0 hmmm ok
18:06 spekki0 I'll try catch the error better then and just skip over those records
18:06 spekki0 at least it's not a bioperl issue
18:07 pyrimidine I'll check it out, JIC
18:07 * rbuels assigns deafferret the ticket to implement Bio::NCBI::Fixer
18:12 pyrimidine spekki0: works for me: http://pastebin.com/nhnDNUed
18:12 spekki0 kk lemme look
18:17 spekki0 hmm ok well i'll play around with it some more only thing different is im using epost before with -keep_hisotries = 1 and passing the histories object to efetch
18:18 pyrimidine that might explain some of it
18:18 pyrimidine spekki0: if this persists, can you post a bug report with some code?
18:19 spekki0 I will, I'll try a couple more things later and come back and let you know if your still around. Im going for lunch, thanks for helping with where the problem might be.
18:20 pyrimidine np
18:24 pyrimidine http://www.spiegel.de/internatio​nal/world/0,1518,709174,00.html
18:24 pyrimidine got this from #bioinformatics
19:04 spekki0 k, so im trying to figure out my problem a bit more and im running into more issues with epost here is my code http://codepad.org/dqob0iGw and it ends up not downloading and basically braeking on my $count = $factory->get_count; am i missing something?
19:13 pyrimidine spekki0: works for me with a few changes
19:13 pyrimidine http://codepad.org/tLvLWX1N
19:13 pyrimidine two problems
19:13 pyrimidine 1) epost doesn't return a count
19:13 spekki0 ah that would explain that part
19:14 pyrimidine 2) get_Response(-file => 'output.txt' )
19:14 pyrimidine that should be a file name, not a file handle
19:14 spekki0 oh, can it use a fh?
19:14 spekki0 just curious
19:14 pyrimidine IIRC it will accept a file handle using -fh
19:14 * pyrimidine checking
19:15 pyrimidine well, sort of
19:15 pyrimidine you can pass a callback that prints to a file handle
19:16 spekki0 yeah i just saw that, -cb => sub {my ($data) = @_; print $out $data}
19:16 pyrimidine yes
19:16 pyrimidine these are just delegated to the HTTP::Response
19:17 pyrimidine it is possible to add a -fh option, though, which will just do the above
19:17 spekki0 cool
19:23 spekki0 hmm ok so here is my modified code http://codepad.org/2PaU6o2u still gives me a bit weird results the first gi is good the second one has a record that should be removed but it still ends up downloading the record. Is this a bug on NCBI's part or something else?
19:25 pyrimidine spekki0: my guess is yes, that is a bug on NCBI's end.
19:26 spekki0 ok, ill submit a bug report about it at the end of the day, anyone know the email offhand?
19:26 pyrimidine it's actually still in the database (you can click on it via entrez)
19:26 pyrimidine so they may not consider it a bug
19:27 spekki0 hmmm, then how do you deal with it being removed, that's what im wondering?
19:27 pyrimidine in order to check the status you may need to run esummary or somesuch
19:28 spekki0 hmm ok never used those much, ill look at the stuff in the wiki's eutiles cookbook, but any pointers on what exactly to check?
19:28 pyrimidine yes, that's it
19:29 pyrimidine http://codepad.org/z8dt3gdW
19:30 pyrimidine second record: status is suppressed, comment indicates why
19:30 spekki0 thanks dude your a life saver on this one
19:31 pyrimidine np
19:58 CIA-94 bioperl-live: Chris Fields master * r56ba538 / Bio/SeqFeature/Lite.pm : return the first note (scalar context of array, was returning number of notes) - http://bit.ly/cJG4DR
20:58 pyrimidine left #bioperl

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