Camelia, the Perl 6 bug

IRC log for #bioperl, 2011-07-19

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

All times shown according to UTC.

Time Nick Message
00:25 dnewkirk left #bioperl
00:42 ty joined #bioperl
01:05 scottcain left #bioperl
01:07 scottcain joined #bioperl
01:29 scottcain left #bioperl
02:01 ty left #bioperl
02:32 scottcain joined #bioperl
04:32 scottcain left #bioperl
04:54 tiago left #bioperl
04:54 tiago_ joined #bioperl
04:55 tiago_ left #bioperl
04:58 zenman_ joined #bioperl
05:07 zenman_ left #bioperl
05:31 buelsbot left #bioperl
05:45 svaksha left #bioperl
07:24 tiago joined #bioperl
08:40 svaksha joined #bioperl
11:32 carandraug joined #bioperl
14:35 rbuels left #bioperl
16:46 carandraug Hi! Anyone has experience with the Ensembl? I'm reading their documentation but I can't find something that I thought would be quite simple. Given a search term, how can I get a list of gene IDs?
16:58 pyrimidine left #bioperl
17:17 dbolser carandraug: I've been looking at it recently
17:17 dbolser but I didn't come accross searching
17:18 carandraug dbolser, same here. I 'm writing an em-ail to ensembl dev asking how to do it. Should be something really simple
17:18 dbolser asking on the ml++
17:18 rbuels joined #bioperl
17:19 carandraug dbolser, they have an archive of the mailing list but it's not searchable
17:19 dbolser carandraug: what fields are you hoping to search?
17:19 dbolser carandraug: can you trick google into searching it for you?
17:19 dbolser http://www.google.co.uk/webhp?sourceid=chr​ome-instant&ie=UTF-8&ion=1&nor​d=1#sclient=psy&hl=en&nord=1&s​ite=webhp&source=hp&q=site%3Alists.ensembl.org%2Fpipermail%2Fdev%2F%20searc​h&aq=&aqi=&aql=&oq=&pbx=1&fp=46c8377fe9b7​fe43&ion=1&ion=1&bav=on.2,or.r_gc.r_pw.&f​p=46c8377fe9b7fe43&ion=1&biw=1186&bih=744
17:20 dbolser googls urls are so nice
17:20 carandraug dbolser, probably could. The search I want to make would be the same as one would do on the ensembl website
17:20 dbolser http://goo.gl/4dlMR
17:21 dbolser carandraug: what fields are you hoping to search?
17:21 carandraug dbolser, i got it working no problem with EUtilities. make a query, get a list of uids
17:21 carandraug dbolser, what do you mean fields?
17:21 dbolser yah, but the whole NCBI api is based on searching pretty much
17:22 dbolser do you want to search the 'gene name' for example, or the 'gene symbol' etc
17:23 carandraug dbolser, oh! I don't know. The idea is to make a general tool. Today I will be searching hist1h2a*[gene symbol], tomorrow I may search for something else
17:23 dbolser ic
17:23 dbolser carandraug: you'd prolly get quite far with just xrefs
17:24 carandraug dbolser, does that make sense to you? I want the same results one would get when searching on their main page
17:24 dbolser ah, that's a different question entirely
17:25 dbolser just ask that on the emailing list
17:25 dbolser or use LWP to hit their html pages ;-)
17:25 carandraug dbolser, I'm going through the records. Should be something real easy, right?
17:25 dbolser sorry, when you mentioned 'search the ensemble site', I thought you were referring to the google 'site' search thingy
17:26 dbolser carandraug: not necessarily, it depends on how they code their own site wrt their own api
17:26 dbolser they may be too l337 to use the api internally
17:26 dbolser but, yeah, it *should* be exposed via the api
17:27 carandraug dbolser, I hope so
17:28 carandraug dbolser, I can't find any reference how to do it on their tutorials and no reference on the mailing list either
17:28 dbolser rbuels: I was thinking about soaping up ensemble... are there some tricks to use to wrap an existing package full of functions in a soap call? I was thinking somehting like the way moose can be used to wrap a non moose package
17:29 dbolser carandraug: fire up the emacs
17:29 dbolser and send them a message
17:29 carandraug dbolser, already doing that
17:29 dbolser :-)
17:29 rbuels dbolser: i'm not aware of tricks for it, but there may be.  i would ask in #catalyst on irc.perl.org, and #perl on freenode also
17:30 dbolser rbuels: how come it's a catalyst thing?
17:30 rbuels dbolser: there are probably at least some web framework things that will make it easier for you to construct soap endpoints
17:30 dbolser ic
17:30 rbuels dbolser: it's not a catalyst thing, but the people in #catalyst are likely to know
17:30 dbolser ty
17:30 carandraug dbolser, are you a bioperl dev? I noticed that the scripts have their names changed to bp_scriptname during install. Any reason why not to have the files already with bp? I noticed the man pages refer to the scripts without the bp. I forked it and was planning on fix that
17:31 dbolser carandraug: I'm not really a dev, no. rbuels or py<tab> may know (when he comes in)
17:32 dbolser ++ on the planning to fix things
17:32 rbuels carandraug: i'm not sure what the logic is behind having it that way.  pyrimidine is likely to be in the channel pretty soon, he would probably know
17:32 rbuels carandraug: i agree with you that it's silly
17:33 carandraug rbuels, I may fix it after confirming with him then
17:34 dbolser http://www.facebook.com/pages/Th​e-Potato-Genome/224374324267873
17:34 carandraug ahahahah! The potato genome like spotato
17:36 * rbuels also thinks facebook, in general, is silly
17:37 rbuels but that's just cause i'm a crusty old codger.
17:47 dbolser I started a page as an experment
17:50 rbuels dbolser: oh, it's probably a good thing to do
19:23 pyrimidine joined #bioperl
19:35 philsf joined #bioperl
19:35 philsf left #bioperl
19:35 philsf joined #bioperl
19:44 carandraug pyrimidine, earlier today I asked 'I noticed that the scripts have their names changed to bp_scriptname during install. Any reason why not to have the files already with bp? I noticed the man pages refer to the scripts without the bp. I forked it and was planning on fix that'
19:44 carandraug <rbuels> carandraug: i'm not sure what the logic is behind having it that way.  pyrimidine is likely to be in the channel pretty soon, he would probably know
19:44 carandraug <rbuels> carandraug: i agree with you that it's silly
19:53 svaksha left #bioperl
19:54 svaksha joined #bioperl
19:56 svaksha left #bioperl
19:56 svaksha_ joined #bioperl
19:58 svaksha_ left #bioperl
19:59 svaksha joined #bioperl
19:59 svaksha left #bioperl
19:59 svaksha joined #bioperl
20:36 pyrimidine carandraug: yes, it is silly.  I recall there being a reason for this at some point, but I slept since then.  I think it's safe to go ahead and change them.
20:58 philsf left #bioperl
21:13 carandraug pyrimidine, cool. Will do so
21:15 carandraug dbolser, hey. I got an answer from them. Turns out they don't use their own API, they use a search engine (Lucene)
22:29 carandraug left #bioperl
22:32 dbolser ...
22:32 dbolser so close
22:33 * dbolser wonders why they don't include the Lucene search in the api
22:33 dbolser night

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