From: Alex B. <aj...@gm...> - 2015-08-27 17:18:09
|
Hi Celera Community, I'm using celera 8.3 and getting an error message I can't seem to make heads or tails of: ----------------------------------------START Thu Aug 27 08:13:38 2015 /global/common/genepool/jgi/assemblers/celera/8.3/Linux-amd64/bin/overlapStore -u /global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/hp.ovlStore /global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/3-overlapcorrection/hp.erates > /global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/3-overlapcorrection/overlapStore-update-erates.err 2>&1 ----------------------------------------END Thu Aug 27 08:14:17 2015 (39 seconds) ERROR: Failed with signal HUP (1) ================================================================================ runCA failed. ---------------------------------------- Stack trace: at /usr/common/jgi/assemblers/celera/8.3/Linux-amd64/bin/runCA line 1649. main::caFailure('failed to apply the overlap corrections', '/global/homes/a/ajbrandt/bscratch/h/celera_attempt/h...') called at /usr/common/jgi/assemblers/celera/8.3/Linux-amd64/bin/runCA line 4569 main::overlapCorrection() called at /usr/common/jgi/assemblers/celera/8.3/Linux-amd64/bin/runCA line 6557 ---------------------------------------- Last few lines of the relevant log file (/global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/3-overlapcorrection/overlapStore-update-erates.err): ERROR: iNum 291211181897 != orig->ovs.numOverlapsTotal 418401494276 I get that there must be some discrepancy between iNum and the number of total overlaps, but I have no idea how to fix the problem (what steps to rerun, etc). Thanks Alex |
From: Brian W. <th...@gm...> - 2015-08-27 17:31:14
|
That sounds like at least one of the jobs in 3-overlapcorrection failed to finish properly, though I've never seen that happen. What is in (ls -l) the 3-overlapcorrection directory? If there are still *.err or *.out or some other kind of logging files, scan those for errors (they'll probably be at the end of the file). The process here is to recompute overlaps after making base changes in the reads. The expectation is that each old overlap (numOverlapsTotal) will generate a new error rate (iNum), which can be copied into the store. For some reason, there were fewer new error rates than overlaps. The fix is to rerun - maybe manually - the last step ("ovlcorr") that recomputes overlaps. On Thu, Aug 27, 2015 at 1:18 PM, Alex Brandt <aj...@gm...> wrote: > Hi Celera Community, > > I'm using celera 8.3 and getting an error message I can't seem to make > heads or tails of: > > ----------------------------------------START Thu Aug 27 08:13:38 2015 > > /global/common/genepool/jgi/assemblers/celera/8.3/Linux-amd64/bin/overlapStore > -u > /global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/hp.ovlStore > /global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/3-overlapcorrection/hp.erates > > > /global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/3-overlapcorrection/overlapStore-update-erates.err > 2>&1 > > ----------------------------------------END Thu Aug 27 08:14:17 2015 (39 > seconds) > > ERROR: Failed with signal HUP (1) > > > ================================================================================ > > > runCA failed. > > > ---------------------------------------- > > Stack trace: > > > at /usr/common/jgi/assemblers/celera/8.3/Linux-amd64/bin/runCA line 1649. > > main::caFailure('failed to apply the overlap corrections', > '/global/homes/a/ajbrandt/bscratch/h/celera_attempt/h...') called at > /usr/common/jgi/assemblers/celera/8.3/Linux-amd64/bin/runCA line 4569 > > main::overlapCorrection() called at > /usr/common/jgi/assemblers/celera/8.3/Linux-amd64/bin/runCA line 6557 > > > ---------------------------------------- > > Last few lines of the relevant log file > (/global/homes/a/ajbrandt/bscratch/h/celera_attempt/h_parallel/3-overlapcorrection/overlapStore-update-erates.err): > > > ERROR: iNum 291211181897 != orig->ovs.numOverlapsTotal 418401494276 > > > I get that there must be some discrepancy between iNum and the number of > total overlaps, but I have no idea how to fix the problem (what steps to > rerun, etc). > > > Thanks > > Alex > > > ------------------------------------------------------------------------------ > > _______________________________________________ > wgs-assembler-users mailing list > wgs...@li... > https://lists.sourceforge.net/lists/listinfo/wgs-assembler-users > > |