Menu
▾
▴
Re: [wgs-assembler-users] failed to unitig
|
From: Walenz, B. <bw...@jc...> - 2012-12-11 22:45:08
|
Hi Audrey- Very strange. It computed a non-positive overlap length! This could be a corrupt overlap store, but lets hope not. I added some debug code to both the bog and bogart unitiggers, but you’ll need to upgrade to the CVS version to get it. I also made it accept that the overlap is problematic, and just not use it, instead of crashing. It will output warning messages (to unitigger.err) when this happens. To restart, just delete the 4-unitigger directory and run runCA from the newer code base. If there are only a few warnings, you’re probably OK. Losing a few overlaps won’t kill the assembly. If there are lots, overlaps will need to be recomputed to rebuild the store. As for the spec, the only change I’d make is to use bogart instead of bog. I don’t see any dnc* options -- did you clean up the PE/MP confusion in the long-insert Illumina? (http://sourceforge.net/apps/mediawiki/wgs-assembler/index.php?title=Pair_classification_within_Illumina_mate_pair_data) b -- Brian Walenz Senior Software Engineer J. Craig Venter Institute On 12/11/12 11:10 AM, "Audrey Nisole" <aud...@ho...> wrote: Hi, I used runCA7.0 on a large hybrid assembly with 454 and Illumina data. 454 libraries were single and 6kb paired end, and Illumina libraries were short and long-insert Paired End. There's a total of 20 Gb for up to 100 000 000 fragments. Everything went well at the beginning but after completing the different overlap process, the program stopped during the 4-unitigger step with the following error message : #----------------------------------------START Sun Dec 9 13:39:45 2012 #/prg/wgs/7.0/Linux-amd64/bin/overlapStore -u /budworm_wgs_20121120/budworm_20121120.ovlStore /budworm_wgs_20121120/3-overlapcorrection/budworm_20121120.erates> /budworm_wgs_20121120/3-#overlapcorrection/overlapStore-update-erates.err 2>&1 #----------------------------------------END Sun Dec 9 13:58:20 2012 (1115 seconds) #----------------------------------------START Sun Dec 9 13:58:20 2012 #/prg/wgs/7.0/Linux-amd64/bin/buildUnitigs -O /budworm_wgs_20121120/budworm_20121120.ovlStore -G /budworm_wgs_20121120/budworm_20121120.gkpStore -T #/budworm_wgs_20121120/budworm_20121120.tigStore -B 782495 -e 0.03 -E 2.5 -b -m 7 -U -o /budworm_wgs_20121120/4-unitigger/budworm_20121120 > /budworm_wgs_20121120/4-unitigger/unitigger.err #2>&1 #sh: line 1: 41499 Abandon /prg/wgs/7.0/Linux-amd64/bin/buildUnitigs -O /budworm_wgs_20121120/budworm_20121120.ovlStore -G /budworm_wgs_20121120/budworm_20121120.gkpStore -T #/budworm_wgs_20121120/budworm_20121120.tigStore -B 782495 -e 0.03 -E 2.5 -b -m 7 -U -o /budworm_wgs_20121120/4-unitigger/budworm_20121120 > /budworm_wgs_20121120/4-unitigger/unitigger.err 2>&1 #----------------------------------------END Sun Dec 9 18:50:41 2012 (17541 seconds) #ERROR: Failed with signal ABRT (6) #================================================================================ #runCA failed. #---------------------------------------- #Stack trace: #Goto undefined subroutine &Carp::shortmess_real at /usr/lib/perl5/5.10.0/Carp.pm line 41.# I started it again to try to make it recovering from a premature termination but it stopped once again with this message : #at /prg/bin/runCA line 1237. # main::caFailure('failed to unitig', '/budworm_wgs_20121120/4-uniti...') called at /prg/bin/runCA line 4382 # main::unitigger() called at /prg/bin/runCA line 5882 #---------------------------------------- #Last few lines of the relevant log file (/budworm_wgs_20121120/4-unitigger/unitigger.err): #Bad mate threshold = -7 #Error threshold = 0.030 (3.000%) #Error limit = 2.500 errors #sizeof(ufPath) = 24 #FragmentInfo()-- Loading fragment information deleted: 0 active: 10000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 20000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 30000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 40000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 50000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 60000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 70000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 80000000 #FragmentInfo()-- Loading fragment information deleted: 0 active: 90000000 #FragmentInfo()-- Loading fragment information deleted: 0 active:100000000 #FragmentInfo()-- Loaded 100159473 alive fragments, skipped 0 dead fragments. #FragmentInfo()-- Saving fragment information to cache '/budworm_wgs_20121120/4-unitigger/budworm.fragmentInfo' #setLogFile()-- Now logging to '/budworm_wgs_20121120/4-unitigger/budworm.001.bestoverlapgraph-containments.log' #buildUnitigs: AS_BOG_BestOverlapGraph.hh:221: uint32 BestOverlapGraph::olapLength(uint32, uint32, int32, int32): Assertion `bovl > 0' failed. #---------------------------------------- #Failure message: #failed to unitig# I'd like to understand what happen. Is there any particular reason for a failure in the unitig process ? Should I correct some specific settings to avoid that premature termination ? I attached the specfile I used. Thanks for any help about that ! Audrey N. |
