Camelia, the Perl 6 bug

IRC log for #bioperl, 2009-09-18

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

All times shown according to UTC.

Time Nick Message
00:39 ende joined #bioperl
02:07 ilbot2 joined #bioperl
02:07 Topic for #bioperlis now http://www.bioperl.org - You can get help here, but you must be patient.  Paste to http://gist.github.com/.
06:15 kyanardag_ joined #bioperl
06:16 ilbot2 joined #bioperl
06:16 Topic for #bioperlis now http://www.bioperl.org - You can get help here, but you must be patient.  Paste to http://gist.github.com/.
06:26 perl_splut joined #bioperl
08:25 timp joined #bioperl
13:23 faceface joined #bioperl
13:23 faceface rbuels: !!!
13:23 faceface help meeeee
13:23 faceface hello
13:30 faceface MSG: Use of Bio::Assembly::Contig::no_sequences() is deprecated
13:30 faceface so what should I use?
14:43 arareko joined #bioperl
15:01 kyanardag_ joined #bioperl
15:08 deafferret faceface: didja check the wiki and/or svn?
15:08 deafferret bugzilla?
15:09 faceface er... no
15:10 faceface num_sequences?
15:10 deafferret oh. is that in Bio::Assembly::Contig ?
15:11 faceface yes
15:11 faceface no
15:11 faceface wait
15:11 faceface I'm using Bio::Assembly::Contig
15:11 faceface but I don't see num_sequences anywhere
15:11 faceface are you really deaf?
15:12 deafferret no, but my ferret was
15:13 * deafferret starts looking
15:13 faceface oh
15:13 faceface plz look here http://www.bioperl.org/wiki/Modul​e:Bio::Assembly::IO#Example_usage
15:15 deafferret num_sequences() says it replaces no_sequences
15:15 deafferret perldoc Bio::Assembly::Contig
15:15 deafferret so you could update the wiki
15:18 faceface will do
15:18 faceface at the above location, am I confusing assembly with scaffold in the comments?
15:19 deafferret I shouldn't comment on that since I don't understand assembly scaffolds. I'd be guessing.
15:21 faceface right
15:22 faceface ## Get the individual Assembly objects from the file
15:22 faceface ## $assem is a Bio::Assembly::ScaffoldI compliant object
15:22 faceface so I think I should say, get the individual scaffold objects from the file. $scaff is a ...
15:25 faceface Bio::Assembly::Scaffold ... mehod 'id' ... $assembly->id()
15:25 faceface why is that under Scaffold?
15:26 deafferret no clue. I'd just be guessing.  :)
15:27 joris__ joined #bioperl
15:31 joris__ hey, if I wanted to align 2 sequences, and get the % similarity. What Should I do best ? With bl2seq I get sometimes multiple results from fragments: Identities = 58/62 (93%) , Identities = 18/19 (94%), but my input sequences are 600 long ... so I want to have like 1 % for the whole similarity between the 2 seq
15:32 faceface thanks for help... don't feel bad because its seriously garbled in the docs / code... if you didn't know assembly / scaffold / contig / reads very well, it would be uber hard.
15:32 faceface joris__: try sw
15:32 faceface in emboss
15:33 faceface joris__: the 'problem' your seeing is because bl2seq is a local alignment
15:36 faceface and you want a global alignment
15:40 deafferret joris__: clustalw
15:58 faceface OK algo... I have the position of a gap in a seq, and I have the position of an aligned read to the seq. How do I find if the alignment pos encloses the gap?
15:58 faceface er...
15:58 faceface vectors of both
16:01 deafferret can you set up a little github demo?
16:05 faceface sorry, its a dumb question
16:08 deafferret I'm interested in this problem, but don't know how to get started
16:08 deafferret so a github demo would be neato
16:15 joris__ faceface: Thx, Ill try to get that working in Perl, at least on the internet it already gives me the results I expected :)
16:16 faceface cool
16:16 faceface deafferret: awesome
16:16 faceface deafferret: how do I set up a github?
16:16 faceface deafferret: we have some assemblies that we hope to 'patch' with 454 data
16:16 deafferret github.com
16:17 faceface so I'm trying to see which reads map across gaps
16:17 deafferret have you used git?
16:17 faceface no
16:17 deafferret svn? vcs? rcs?
16:17 faceface but its like svn no?
16:17 faceface yes
16:17 deafferret ya  :)
16:17 faceface yes, no, no
16:17 deafferret github.com for a free account, set up repos, ssh keys etc
16:18 * faceface prepares for technology update
16:18 deafferret everybody's doing it.   :D
16:18 faceface the facebook of dev code?
16:19 * faceface wonders if username dan is gone...
16:21 faceface Login has already been taken :-(
16:21 faceface Login is invalid :-(
16:21 faceface Welcome to GitHub!
16:21 deafferret looks like the free plans are 300MB
16:21 faceface I sea
16:22 faceface Project Name?
16:22 faceface Opperation deafferret ?
16:22 deafferret well, I just do this   http://github.com/jhannah/sandbox
16:22 faceface A squid cat tells me no
16:22 faceface http://github.com/danbolser/sandbox
16:23 faceface should I create a sandbox project?
16:23 deafferret yup
16:23 faceface Homepage URL?
16:23 deafferret you can call your repos whatever you want. "sandbox" is my misc-whatever repo
16:23 faceface OK
16:23 deafferret i don't think I have one.
16:23 * faceface tries google
16:24 deafferret heh. you could snag bio.github.com
16:24 faceface wha?
16:25 faceface ever installed mercurial?
16:25 faceface you need to progressively move through 4 other version controll systems to get it
16:26 faceface deafferret: couldn't I just paste you teh code?
16:26 faceface The monsters are putting me off
16:27 deafferret whatever makes you happy
16:28 faceface http://github.com/danbolser/sandbox
16:28 faceface sooo... I'll put my scripts there...
16:28 deafferret faceface++
16:29 deafferret oh, THAT Homepage... laugh
16:30 faceface er...
16:31 faceface ok, files are cp'ed there
16:31 faceface now I git add them?
16:32 faceface or git ... push ... somehting?
16:32 deafferret git add; git commit; git push
16:32 deafferret I'm enjoying your README so far  :)
16:33 faceface its what it told me to do !
16:33 faceface ;-)
16:35 faceface so commit sends it to my local repo, and push pushes it onto the hub?
16:35 deafferret ya
16:35 faceface I heard someone talking about 'working on a plane'
16:35 faceface using git hub to 'commit'
16:35 deafferret git, not github  :)
16:35 faceface er...
16:35 faceface erm..
16:36 deafferret git uses 100% independent repos everywhere
16:36 faceface http://github.com/danbolser/sandbox
16:36 faceface what does that mean?
16:36 deafferret that's your git repository hosted by github
16:36 deafferret but you can host your own git repos anywhere you want
16:36 deafferret and let people push/pull
16:36 faceface what does "git uses 100% independent repos everywhere" mean?
16:36 faceface via http?
16:37 deafferret In subversion you can't commit unless you can talk to the server.
16:37 faceface now the squid cate is crying!
16:37 deafferret In git that is never the case. The full repo is always local.
16:37 faceface right
16:37 faceface and you push it out to ... server?
16:37 deafferret so you can commit 50 times on a flight, then when you land, push your commits to 7 different servers
16:38 faceface why?
16:38 faceface ;-)
16:38 deafferret so you can code on a plane! or a desert island!
16:38 deafferret so when the aliens attack your coding is not interrupted
16:38 deafferret (unless you need to connect to a remote database. then you're still screwed.)
16:39 deafferret (which is why I don't use git for $work)
16:39 faceface can I put a git hub on google maps?
16:39 faceface street view
16:39 * deafferret steals some of faceface's drugs
16:40 faceface lol
16:40 deafferret hey, looky there. git pull pulled some gaps  :)
16:40 deafferret what's .plx?
16:40 * deafferret checks the README
16:40 faceface the gaps are gff, and the 'mapped reads' were created using ace2tab from contigs.ace
16:40 faceface there is no readme ;-)
16:40 deafferret I'm staring right at it! All 0 bytes!
16:41 faceface why did firefox the javascript break?
16:41 faceface 3.0.xx
16:41 deafferret so what should I be poking with this here stick?
16:43 faceface git commit -m 'Wrote the README'
16:43 faceface On branch master Changed but not updated:  (use "git add <file>..." to update what will be committed)
16:43 * faceface tries pushing
16:44 faceface Everything up-to-date
16:44 deafferret what are you trying to do?
16:44 faceface deafferret: you said "I'm interested in this problem"
16:44 faceface trying to update the readme
16:44 deafferret Indeed, sir.
16:45 deafferret vi README; git add README; git commit; git push;
16:45 faceface deafferret: the problem of intersecting ranges, or the problem of patching assemblies using 454 reads?
16:45 faceface oh... howcome I need to add it?
16:45 faceface readme added
16:45 deafferret or you can   git commit -a   which says "commit everything that changed"
16:46 faceface ok
16:46 faceface so can you grab that readme and change it now?
16:46 deafferret I'm not sure why they do that, exactly. there's lots of fancy git commands for splitting out what-changes-to-what-files​-should-go-in-what-commits
16:46 deafferret so commits are super-super clean for others to review
16:47 deafferret which also causes the whole rebasing holy war
16:47 faceface oh
16:47 faceface you mean svn -> git?
16:47 deafferret no, I was talking solely about git there
16:47 deafferret rebasing is a git thing of much heated debate
16:48 deafferret and hatred
16:48 deafferret By default the github way of collaborating with me is that I commit changes to my clone and then I send you a "pull request" -- you then pull my changes into your repo
16:49 deafferret which is all point-and-click in github
16:50 * deafferret rubs his hands together, excitedly
16:50 deafferret so you don't have an immediate BioPerl problem, this is just the state of your project right now?
16:50 * deafferret squeals with glee
16:51 deafferret what's .plx? why not .pl?
16:52 faceface deafferret: I have a kind of bioperl problem in that I don't know how to find the aligned region of my read
16:52 faceface plx is better ;-)
16:52 * faceface has no good answer
16:52 faceface ... the first thing I read about perl said "use plx, pl means perl library"
16:52 faceface I never looked back
16:53 faceface see mail on mailing lsit
16:53 deafferret wow. what year was that?
16:53 faceface er... even the book said "but nobody does that"
16:54 deafferret hmmm so Bio::Assembly::IO doesn't like this ACE file... no next_assembly()
16:54 deafferret ace2tab.plx 454Contigs.ace.tab   right?
16:54 faceface deafferret: no sorry
16:55 faceface ace2tab.plx 454Contigs.ace
16:55 faceface but the ace is large
16:55 faceface where do you want it?
16:55 faceface oh wait, I think I have a smaller one here somewhere
16:55 deafferret how large? can you head -100000 it or something?
16:55 deafferret BioPerl t/data might have some too
16:55 deafferret that
16:55 deafferret thanks
16:56 faceface right, but the gaps won't match ... still, you can play with it
16:56 * deafferret rubs rbuels' chin
16:57 faceface 3.8G
16:57 * deafferret chokes
16:57 faceface lol
16:57 deafferret stupid biology. being all complex and stuff
16:58 faceface there is a split version... 57,000 files in one dir borks though
16:58 deafferret so ace2tab.plx created 454Contigs.ace.tab ?
16:58 faceface yes
16:58 faceface I'll send you an ace to try it on...
16:58 deafferret i see. so I'll jump straight to gaps_closed.plx
16:58 deafferret phone ringing
16:58 faceface then 454Contigs.ace.tab is combined with the gaps.gff to see which reads 'span' gaps
16:59 faceface hide!
16:59 faceface 32M?
16:59 faceface OK, this one is 6.0M
17:06 faceface pushed
17:07 faceface you should be able to process the .ace into a .ace.tab using ace2tab.plx, then look at the 'gap spanning reads' using gaps_closed.plx <- which is very 'work in progress'
17:10 faceface deafferret: g2g
17:24 deafferret faceface: sorry, long call. back in a min.   :)
17:25 deafferret come back whenever you have time again  :)
17:28 spvensko joined #bioperl
17:29 spvensko hi, does anyone here know if bioperl has any sort of module that can be used to connect to wormbase and extract information from it? i've looked in the wiki and can't seem to find much
17:34 deafferret spvensko: wormbase-specific info? like what? do they have APIs for retrieving their data?
17:35 spvensko i am trying to get nucleosome coverage information for the genome but i can't even get those to show up in the browser
17:35 deafferret Oh, it's a GMOD so maybe many GMOD APIs applyu
17:35 spvensko i haven't look into what kind of APIs they have, i'm still fairly new to all of this and didn't think to check
17:36 deafferret hmm... if that info isn't in their web interface at all you should probably email them and ask if it exists?
17:39 spvensko probably a good idea
17:41 deafferret have you looked in here?  http://www.wormbase.org/biomart/martview
17:58 spvensko no but it looks like that's what i need, thank you
18:06 deafferret faceface: ok, I think I understand this software now. What's crude about it? Seems find given the dataset?
18:07 deafferret faceface: thanks for sharing it with me  :)  if I can help I'd love to try
19:35 meteorswarm joined #bioperl
19:42 spvensko left #bioperl
19:44 dnewkirk joined #bioperl
19:53 meteorswarm I'm starting a project that's attempting to integrate some gene expression data with metabolic pathways from TAIR (araCyc).  Are there any bioperl packages that would help me?  I can't find anything on the wiki.
20:03 faceface meteorswarm: not that I know of
20:03 faceface cytoscape does something similar I belive
20:03 faceface search wiki for cytoscape?
20:04 faceface spvensko... in bioperl...
20:04 faceface over in #bioinformatics he's all about hte python ;-P
20:04 faceface the
20:05 faceface deafferret: your very welcom :D
20:05 meteorswarm faceface: looking up cytoscape
20:05 faceface deafferret: the crudity comes from the lack of 'propper' analysis ... i.e. I need to check the depth of the coverage in the spanning reagion / the quality of the alignment either side of the gap
20:05 faceface but thanks for the encouragement :-)
20:05 deafferret faceface: I have yet to find a killer app in BioPython
20:06 faceface I'm often quite negative
20:06 faceface deafferret: lets face it, biopython is just a bioperl wannabe
20:06 faceface ;-)
20:06 meteorswarm faceface: bioperl wiki turns up nothing on cytoscape, should I be looking somewhere else?
20:06 faceface meteorswarm: try teh webs
20:06 faceface cytoscape is a stand alone package, but I don't really know how it works, so I wondered if there were some perl wrappers
20:07 faceface iirc it does something similar to what you suggested.
20:07 faceface meteorswarm: I never did microarray analysis, so I have never looked for any tools in bioperl
20:07 faceface however, I know there are a lot of such tools in BioC
20:07 faceface (which is BioR ...)
20:08 faceface I wonder why they didn't call it BioR
20:08 meteorswarm faceface: I'll take a look, thanks.  The cytoscape thing looks pretty fancy.  We're not really doing microarray stuff, we've done a bunch of transcriptome sequencing and I'm working from that data, but I guess it's sort of equivalent
20:09 faceface meteorswarm: right
20:09 faceface meteorswarm: you could look at chado
20:10 kyanardag_ joined #bioperl
20:11 meteorswarm faceface: I will.  I was actually just looking for a parser for now :/  I've greatful for the leads though, they look useful.
20:11 meteorswarm I suppose it wouldn't be terrible to parse as it stands
20:11 faceface parser for txomics? what platform?
20:12 faceface meteorswarm: NGS people need to stick together ;-)
20:12 faceface seen seqanswers?
20:12 meteorswarm I'm not grokking your abbreviations
20:12 faceface heh
20:12 faceface txomics = transcriptome sequencing
20:13 faceface NGS = Next Generation Sequencing
20:13 meteorswarm I'm actually taking the transcriptome and trying to compute the metabolome, which is the ultimate goal.
20:13 faceface platform = Illumina-Solexa / 454 / Solid?
20:13 faceface cool
20:14 meteorswarm ahhh
20:14 meteorswarm it's mostly a year or so-old solexa sequences; 36bp
20:14 meteorswarm they're a tad short :(
20:14 faceface bioperl can handle fastq, but be sure to match versions or your qualities will be out
20:15 faceface meteorswarm: for txs it don't matter
20:15 faceface txs = very lazy way to type txomics
20:15 faceface transcript omics ... man thats a lot of characters!
20:15 faceface t9omics
20:16 meteorswarm faceface: dealing with the sequences has already been done; I've been working with the data for other projects all summer, so that's not really an issue.  It's almost trivial to tally the number of reads per gene since we have gene models already.
20:16 faceface meteorswarm: you need a mapper such as bowtie ... and a reference genome of course
20:16 meteorswarm (Working on Glycine)
20:16 faceface right
20:16 faceface ok
20:16 meteorswarm faceface: they're already mapped and everything
20:16 faceface cool... so you just need to link it to metabolic pathways
20:16 faceface see kegg
20:16 faceface or reactome
20:17 dnewkirk Cytoscape is very useful, if a bit fancy. The DAVID bioinformatics site might be useful for a first glance as well.
20:17 faceface reactome = kegg + hype
20:17 meteorswarm what I need is a way to get machine readable metabolic pathway data and use it; I have the data (I think) from the aracyc website
20:17 faceface meteorswarm: cool
20:18 faceface kegg has a flat file database called enzyme
20:18 faceface or ligand sorry
20:18 faceface also brenda is usefull
20:18 meteorswarm faceface: well, again, I have the database, or at least what claims to be the database, for Arabidopsis, I just don't have a good way to make sense of it
20:18 faceface do any other fields of science argue so much over concept space as biologists?
20:19 faceface how many reaction databases do we need?
20:19 meteorswarm faceface: well, they're different in different, you know, kingdoms
20:19 faceface meteorswarm: kegg, reactome, brenda, they all try to be species agnostic
20:19 faceface meteorswarm: sounds like you want to do something like ordinary differential equation simulation
20:20 meteorswarm faceface: pretty much
20:20 faceface its been around for a while, but these days they call it systems biology
20:20 meteorswarm faceface: the difference, I think, is the scale, since we're trying to do it over the entire transcriptome
20:20 faceface you need some-ML ... I forget its name... SBML?
20:20 faceface right
20:21 faceface but how critically will the results depend on uncertanty in enzymatic rates?
20:21 faceface and other unknowns such as transport coeficients?
20:21 meteorswarm faceface: well, sure, that's a problem
20:21 faceface I think you can get some crude results easily, but convincing yourself that they are acurate... could be tough
20:22 faceface if you want qualitative answers, I think it should be OK
20:22 faceface well... good luck :-)
20:22 meteorswarm faceface: we also have transcriptomes from plants in different conditions, so comparing those would be pretty easy - finding which pathways are over/under expressed
20:22 faceface meteorswarm: right
20:23 faceface just linking that at the metabolic level should be interesting.
20:23 meteorswarm that's the idea :D
20:23 meteorswarm heading out, thanks for the help
20:55 * deafferret throws up a little
20:55 deafferret "a set of Excel macros for use in linking functional gene class associations to a given set of genes"
20:55 deafferret http://sigterms.sourceforge.net/
21:21 faceface love that macro
23:03 faceface deafferret: now I'm 'titrating' my 454 data againt the N50 of the 'patched' assemblies :-)
23:06 faceface the idea is to show that we have a) enough 454 data or b) not nearly enough 454 data.
23:58 ptl joined #bioperl

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