Open GoogleCodeExporter opened 9 years ago
Sorry the *.sh script did not get attached. Perhaps change the file extension
to TXT or wrap it in a ZIP file and repost.
What was the reason to suspect the script was the problem? It seems like
SAMtools is generating a floating point runtime exception, no?
Original comment by iwaterp...@gmail.com
on 20 Jul 2011 at 11:42
I attach here the zipped script. I was asking bc a line of the script was
recalled..so I interpreted that was a script problem. This "floating point
runtime exception" means that is a problem of samtools? I will try also with
the new samtools build.
fed
Original comment by federica...@gmail.com
on 20 Jul 2011 at 11:51
Attachments:
I ran this module also with an upgraded version of samtools, but I have back
the same error.
Original comment by federica...@gmail.com
on 21 Jul 2011 at 9:58
Hi,
I haven't had any feedback from andrew on this issue, let me know if on your
server you are experiencing the same issues. I re-run on fgene2 also, but I got
back the same floating error(see attached).
Even using command line:
/applications/samtools-0.1.7_x86_64-linux/samtools calmd -b
/projects/pipelineCache/pipeline/2011August19_10h51m17s724ms/FixMateInformation_
1.Output-1
/projects1/Reference_genomes/hg18_ensembl/gatk-canonical/gatk-hg18_ensembl.fa
/projects/pipelineCache/pipeline/2011August19_10h51m17s724ms/SamToolsCamldMDtag_
1.OutputNo-DuplicatesBAMfile-1.bam
Floating exception
Fed
Original comment by federica...@gmail.com
on 19 Aug 2011 at 8:13
Attachments:
Sorry for not responding. We are experiencing the same floating point error on
our servers. Still looking into solutions to this.
Sam
Original comment by shobe...@gmail.com
on 22 Aug 2011 at 7:12
Thanks,
Fed
Original comment by federica...@gmail.com
on 26 Aug 2011 at 8:00
The solution to this problem is that you need full permission to the reference
file. I cannot change the permissions to this file so instead, I created a
symbolic link to the it and used the link as my data source input.
This is the command for symbolic linkage: "ln -s
/projects1/Reference_genomes/hg18_ensembl/gatk-canonical/gatk-hg18_ensembl.fa
/projects1/test_pipeline/NEW_ORGANIZATION/Basic_QC/gatk-hg18_ensembl.fa"
I then put
/projects1/test_pipeline/NEW_ORGANIZATION/Basic_QC/gatk-hg18_ensembl.fa in the
data source.
I modified the workflow significantly to make it simpler and attached it here.
I could even add a module that does the linking automatically, or you can just
make sure you have full permissions to the reference file. Whatever you want.
Sam
Original comment by shobe...@gmail.com
on 14 Sep 2011 at 9:46
Attachments:
More accurately, you need write permission to the directory that the reference
file is in. You don't need full permission to the reference file. This is why
symbolically linking the reference file to a folder you can write to fixes the
problem.
Original comment by shobe...@gmail.com
on 14 Sep 2011 at 10:42
Oh Sam,
thank you so much! I'm gonna test it right away!
fed
Original comment by federica...@gmail.com
on 15 Sep 2011 at 5:11
Hi,
what does it mean when a run is backlogged? is it due to any errors? It's
happening to me for both the QC modules..
fed
Original comment by federica...@gmail.com
on 15 Sep 2011 at 5:26
Attachments:
It just means the server is very busy and cannot accept another job. Although,
you may still be able to run other jobs with smaller memory requirements.
Original comment by shobe...@gmail.com
on 15 Sep 2011 at 6:12
ok, tks a lot I'll shift to any other fgenes
Original comment by federica...@gmail.com
on 15 Sep 2011 at 6:16
Hi,
the run was perfect until the indexing. The output file until then are perfect:
fgene3 [/usr/pl_cache/pipeline/2011September16_14h21m09s250ms] ll
total 19648
-rw-rw-rw- 1 pipeline pipeline 6738598 Sep 16 14:21
SamToolsRemoveDuplicates_1.OutputNo-DuplicatesBAMfile-1.bam
-rw-rw-rw- 1 pipeline pipeline 6592761 Sep 16 14:21
SamToolsSort_1.OutputSortedBAMfile-1.bam
-rw-rw-rw- 1 pipeline pipeline 6738662 Sep 16 14:21
SamToolsView_1.OutputBAMfile-1.bam
drwxrwxrwx 2 pipeline pipeline 4096 Sep 16 16:08 streams
But looking to the indexing process even if it is green, I have the error that
it cannot open the BAM (see pic). If I open up che error logs of the output
folders i have the error saying that it cannot copy the .bai file BUT the .bai
file (as you can see) has not been produced. Note that I have the permissions
on the output folders I have set.
Fed
Original comment by federica...@gmail.com
on 17 Sep 2011 at 12:25
Attachments:
Hi,
the run was perfect until the indexing. The output file until then are perfect:
fgene3 [/usr/pl_cache/pipeline/2011September16_14h21m09s250ms] ll
total 19648
-rw-rw-rw- 1 pipeline pipeline 6738598 Sep 16 14:21
SamToolsRemoveDuplicates_1.OutputNo-DuplicatesBAMfile-1.bam
-rw-rw-rw- 1 pipeline pipeline 6592761 Sep 16 14:21
SamToolsSort_1.OutputSortedBAMfile-1.bam
-rw-rw-rw- 1 pipeline pipeline 6738662 Sep 16 14:21
SamToolsView_1.OutputBAMfile-1.bam
drwxrwxrwx 2 pipeline pipeline 4096 Sep 16 16:08 streams
But looking to the indexing process even if it is green, I have the error that
it cannot open the BAM (see pic). If I open up che error logs of the output
folders i have the error saying that it cannot copy the .bai file BUT the .bai
file (as you can see) has not been produced. Note that I have the permissions
on the output folders I have set.
Fed
Original comment by federica...@gmail.com
on 17 Sep 2011 at 12:26
Attachments:
It looks like the Calmd module outputs the BAM file to the standard output.
Thus, it isn't passing any file to the Index module. I wrote a wrapper script
to redirect stdout to the output parameter and put it here:
/projects1/samtools_calmd.sh. You can copy it wherever you want and use the
path to that script as the executable in the calmd module.
Sam
Original comment by shobe...@gmail.com
on 17 Sep 2011 at 3:02
Use the same script for the calmd module in Basic QC2.
Original comment by shobe...@gmail.com
on 17 Sep 2011 at 3:02
ok, tks running now.
Fed
Original comment by federica...@gmail.com
on 17 Sep 2011 at 3:15
Oh yes, BasicQC1 went ok!
Original comment by federica...@gmail.com
on 17 Sep 2011 at 4:25
Original issue reported on code.google.com by
federica...@gmail.com
on 20 Jul 2011 at 6:49Attachments: