Camelia, the Perl 6 bug

IRC log for #bioperl, 2009-04-14

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

All times shown according to UTC.

Time Nick Message
05:11 ptl sigh
06:03 cotto left #bioperl
06:33 tarbo2 joined #bioperl
09:40 philsf joined #bioperl
10:03 majensen joined #bioperl
10:04 majensen left #bioperl
14:00 deafferret ptl: can you be more specific?  :)
14:29 pyrimidine joined #bioperl
14:30 oss joined #bioperl
14:30 oss Hi!
14:30 * deafferret waves
14:30 pyrimidine Hi right back at you
14:31 * pyrimidine waving back
14:32 oss Hello~ could you help me debug my scripts by bioperl?
14:33 pyrimidine possibly
14:33 pyrimidine do we have an active pastebot?
14:33 pyrimidine pastebot?
14:33 oss I'll use the pastebin, thanks very much.
14:34 pyrimidine well, the pastebin doesn't have #bioperl listed
14:34 oss http://sial.org/pbot/?
14:34 pyrimidine yup
14:35 oss sorry, which channel?
14:36 pyrimidine You can paste unannounced, but it would be nice to get a pastebot going
14:37 pyrimidine just c&p into the block with 'none', and then copy and paste the resulting link
14:37 pyrimidine http://sial.org/pbot/35976
14:37 pyrimidine for instance
14:38 oss http://sial.org/pbot/35977
14:38 oss Thanks for help :)
14:40 pyrimidine what is the problem?
14:40 pyrimidine (beyond "it doesn't work")
14:41 oss What I wrote is a scanner to get the sequences bu using NCBI BLAST output format 8.
14:41 oss But when I run my script I get nothing, and I don't know where's the problems because I didn't get error messages.
14:42 pyrimidine we have a parser for that output (Bio::SearchIO::blasttable)
14:43 oss I guess the BLAST parse I wrote is okay.
14:43 pyrimidine It works.but you're doing a LOT of redundant work
14:43 oss Because the format 8 is tab-delimitaed file.
14:43 oss For instance?
14:43 pyrimidine lots of unnec. splits
14:43 deafferret use the perl debugger or stick print statements in your program to see if it's even opening the file
14:43 pyrimidine yes
14:44 pyrimidine start simply
14:44 pyrimidine Also,
14:44 deafferret are your print statements hitting at all?
14:44 oss I will paste the data I use and could you help me with them?
14:44 pyrimidine you're reloading the sequence stream on each iteration
14:45 pyrimidine seems a better way would be to use index the sequences (Bio::Index) and pull out what you need from that)
14:45 tarbo2 joined #bioperl
14:45 * deafferret has a meeting in 16 minutes
14:46 oss http://sial.org/pbot/35978  <- the first arg
14:47 oss http://sial.org/pbot/35979 <- the second srg
14:47 oss *arg
14:48 oss ./myscript.pl <first> <second>
14:49 * pyrimidine looking
14:50 oss thanks...
14:53 pyrimidine my pseudocode for doing this is...
14:53 pyrimidine # index the sequence file using Bio::Index::Fasta or Bio::DB::Fasta
14:53 pyrimidine # parse the BLAST file ()
14:53 pyrimidine # pull out the sequence you need using the ID from the file
14:53 pyrimidine # grab subseq
14:53 pyrimidine the BLAST file could be parsed as you have it or using SearchIO
14:54 pyrimidine Just to note, it will be much faster to use a preindexed sequence database to retrieve your data vs. rerunning the SeqIO parse
14:54 pyrimidine over and over for each line...
14:56 oss but why I cannot get result?
15:00 oss anyway, could you point out the problems I have at this program?
15:00 pyrimidine you have an off-by-one error with you indexing
15:00 pyrimidine *your
15:00 pyrimidine positions in an array start with 0, not 1
15:00 pyrimidine my $mirna_name = (split /\s+/, $_)[0];
15:00 pyrimidine my $genome_name = (split /\s+/, $_)[1];
15:00 pyrimidine my $gms = (split /\s+/, $_)[8];
15:00 pyrimidine my $gme = (split /\s+/, $_)[9];
15:00 oss I see. oh.......
15:02 oss but I still cannot acquire the subseqs. :(
15:04 pyrimidine this seems to be working for me
15:04 oss you get the result?
15:06 pyrimidine well, I get output...
15:06 pyrimidine http://sial.org/pbot/35980
15:06 pyrimidine though it's a bit mangled...
15:08 oss but I still get nothing.
15:08 oss :(
15:08 oss you modify codes?
15:09 torbjorn joined #bioperl
15:09 pyrimidine nope
15:09 pyrimidine check your line endings
15:10 pyrimidine if you are running off UNIX and use windows line endings (or vice versa) you'll have issues with chomp/split
15:11 tarbo2 joined #bioperl
15:13 oss alright, I'll try.
15:14 oss Oh, no results.
15:15 oss They are unix endings rotally.
15:15 oss totally.
15:16 tarbo2 joined #bioperl
15:16 oss what kinds of problem could happen?
15:20 oss hello?
15:20 pyrimidine patience
15:20 oss sorry.....
15:32 oss Does open fail silently ?
15:34 pyrimidine no
15:36 oss so what do you think?
15:36 pyrimidine I think you need to break your script apart to work out the bug.  It's appears to be a logic issue.
15:37 oss but you didn't modify my codes and run well. Could you paste the script you used now?
15:39 pyrimidine Like I said, you need to test one task out at a time, then assemble them together.  You're doing too much.
15:39 pyrimidine test the parser: http://sial.org/pbot/35981
15:39 pyrimidine for instance (that gives me 'ath-miR157a L1 6-17')
15:41 pyrimidine Then run the iteration through each window size to make sure you are getting the right start/end
15:41 oss I guess the problem is quite odd. Because as you said you run well so I don't suppose that there's something wrong with codes structures, right?
15:42 pyrimidine probably doesn't hurt to check the sequence
15:42 oss Are we using the same sequences?
15:42 pyrimidine yes
15:43 oss so the sequences should be okay. :(
15:43 oss disappointed.
15:45 oss do you have any other good idea?
15:48 pyrimidine not really, but it's definitely a logic issue centered around start/end
15:49 pyrimidine probably has something to do with the way toggle_start/end is calculated
15:49 pyrimidine I have to go to a meeting, I can backlog if you work it out'
15:49 pyrimidine afk
16:21 oss hello?
17:35 rbuels joined #bioperl
18:18 ende joined #bioperl
18:18 krunchky joined #bioperl
18:21 rbuels_ joined #bioperl
20:39 philsf joined #bioperl
21:54 philsf_ joined #bioperl
22:12 philsf joined #bioperl
22:50 ende joined #bioperl
23:54 philsf joined #bioperl

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