Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-05-31

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

All times shown according to UTC.

Time Nick Message
06:28 bag_ joined #bioperl
08:31 vinnana joined #bioperl
10:01 faceface hello vinnana
10:33 CIA-66 joined #bioperl
10:53 vinnana hi faceface
11:10 cawiss joined #bioperl
12:35 brandi joined #bioperl
13:09 brandi left #bioperl
14:35 brandi joined #bioperl
14:35 brandi left #bioperl
14:48 brandi joined #bioperl
14:48 brandi left #bioperl
15:54 spekki01 I'm trying to do home hmmer3 parsing using the bioperl-hmmer3 (http://github.com/bioperl/bioperl-hmmer3) is there any doc's for this kicking around somewhere?
16:08 dnewkirk just the API docs on the site
16:24 kblin spekki01: they work just like the hmmer parser
16:25 kblin just that you specify format => "hmmer3" instead of format => "hmmer"
16:26 kblin however you currently need to copy over the files from bioperl-hmmer3 to the bioperl-live distribution to get it to work
16:39 spekki01 cool, yeah i just finished coppying all the files into there proper directories and ran the test files they have included and it seemed to work, and ill look up the hmm parser on the wiki site
16:44 kblin have fun :)
16:45 kblin and don't hesitate to ping me if you run into problems
17:01 spekki01 kblin: with hmmer3 theres the option of outputting hmmsearch results in a simpler stripped down table output, can the parser parse those files as well?
17:04 kblin er, no idea
17:04 spekki01 kk
17:05 kblin it's the first time I heard of that
17:26 spekki01 not exactly bioperl related but in my hmmsearch result file i have 3492 targets reported over threshold, but then midway between these i have this line saying inclusion threshold. what exactly is the difference between threshehold and inclusion threshold?
17:54 kblin is it midway?
19:17 brandi joined #bioperl
19:21 brandi left #bioperl
19:55 spekki01 with hmmer3 parsing is there a built in way to get the gi number only out of a call to ->name which gives gi|109087045|ref|XP_001094800.1|? or should i just do some text matching and pull it out.
19:55 kblin that really depends on your input sequences
19:56 kblin I assume the FASTA files you used had this header
19:56 spekki01 lemme double check but i think so
19:56 kblin I don't think Bio::SearchIO does anything special with the name of the sequence
19:57 spekki01 yeah the fasta db file has the sequence names all in that format
19:58 spekki01 ill just make my own string parser, tx, was just wondering if there was somethign built in.
19:58 kblin I think the Bio::SearchIO function just returns whatever is in the parsed format, without any post-processing
19:59 kblin and it's likely that hmmer just puts whatever is in the fasta header as the name field as well
19:59 kblin that's what I'd do in a fasta-parser :)
19:59 kblin have you tried that simple hmmsearch format?
20:00 * deafferret waves
20:00 kblin hi deafferret
20:00 * deafferret still knows nothing about hmmer formats  :/
20:01 spekki01 i played around a bit but it seemed like it wasnt working
20:02 spekki01 I'll try it more in depth later
20:02 spekki01 but it would be much nicer to parse through in my opinion
20:03 kblin spekki01: do you happen to have one of these files with only a handful of hits?
20:03 spekki01 well if by handful you mean around 3500 hits, then sure
20:04 kblin hehe, I was thinking about < 20 or so :)
20:04 spekki01 easy enough to just cut out most of the records anyway i can give you a 20 record one if you like
20:04 spekki01 its in a nice easy format to play with
20:05 kblin there's no stats on the end of the file?
20:05 spekki01 nope
20:07 spekki01 lemme post it somewhere for you
20:07 kblin thanks
20:09 spekki01 easier yet ill send it through here lol
20:10 spekki01 always forget about file transfers in irc
20:11 kblin I seem to have missed that one, or you didn't send a file :)
20:11 spekki01 see it now?
20:12 kblin yeah, but you're not sending, I guess some firewall on the path is blocking
20:12 spekki01 hmm thats dumb probably on my end
20:12 spekki01 sry one sec
20:12 kblin however, given it's just 7kb, can you send it to kai.blin(at)biotech.uni-tuebingen.de
20:13 spekki01 coming right up
20:13 kblin thanks
20:20 spekki01 omfg i hate my workstation sometimes
20:46 spekki01 #perl
20:46 spekki01 sry
21:17 kblin spekki01: ok, got the file (yay, graylisting), thanks
21:18 spekki01 np
21:19 kblin from the first look, that's just what's in the normal hmmsearch output, minus all the headers
21:20 spekki01 yeah
21:20 kblin I'm not sure if that'll make for easier or harder parsing
21:20 kblin but it sure makes another interesting test case :)
21:21 spekki01 it would be nice if it was able to parse through that type of file, but that's just gold plating
21:22 kblin the interesting question is what to put into all the fields that are usually filled with the information from the header
21:38 bag_ joined #bioperl

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