Alignment of paired-end reads with gnumaps
39 views
Skip to first unread message
dpry...@gmail.com
Dec 10, 2013, 11:39:30 AM12/10/13
to gnumap...@googlegroups.com
Hi all,
I'm attempting to align some bisulfite converted test reads with the gnumaps_bs package (the most recent version, downloaded yesterday). I'm submitting the job to the cluster (I'm just testing on 3 nodes) as follows (our cluster uses slurm, but you can mentally substitute "mpiexec -N 3" if you prefer):
srun $WORK/gnumaps/scripts/gnumaps.pl --genome $WORK/indexes/genome.fa --pair_1 $WORK/Sequences/test_1.fastq --pair_2 $WORK/Sequences/test_2.fastq
--lib_type wt1 --read_type dna --phred_offset 33 --nt_conv bs --num_threads 12 --outdir $WORK/Alignments/ --skip_blat 1 --pileup 1
$WORK is just shorthand for mountpoint shared by the various nodes. Alignment seems to complete properly (e.g. $WORK/Alignments/test_1/aligned/gnu/gen/watson/ contains 6 appropriately sized SAM files each having what appear to be reasonable alignments), but whatever subsequent processing step gnumaps performs seem to not complete successfully. What I suspect is the relevant portion of the output is as follows (starting with the end of the initial alignment portion):
A similar set of errors is issued for every chromosome and, indeed, those files don't exist (the directory skeleton is created, but never populated with files). I using a multifasta file as input, so might these be causing the behaviour or have I just incorrectly specified an option?
An unrelated question: Is there a way (as with direct use of gnumap) to simply save the hash to a file so it only needs to be created once. Recreating it for every sample that might need to be processed seems rather repetitive.
Thanks much
I'm attempting to align some bisulfite converted test reads with the gnumaps_bs package (the most recent version, downloaded yesterday). I'm submitting the job to the cluster (I'm just testing on 3 nodes) as follows (our cluster uses slurm, but you can mentally substitute "mpiexec -N 3" if you prefer):
srun $WORK/gnumaps/scripts/gnumaps.pl --genome $WORK/indexes/genome.fa --pair_1 $WORK/Sequences/test_1.fastq --pair_2 $WORK/Sequences/test_2.fastq
--lib_type wt1 --read_type dna --phred_offset 33 --nt_conv bs --num_threads 12 --outdir $WORK/Alignments/ --skip_blat 1 --pileup 1
$WORK is just shorthand for mountpoint shared by the various nodes. Alignment seems to complete properly (e.g. $WORK/Alignments/test_1/aligned/gnu/gen/watson/ contains 6 appropriately sized SAM files each having what appear to be reasonable alignments), but whatever subsequent processing step gnumaps performs seem to not complete successfully. What I suspect is the relevant portion of the output is as follows (starting with the end of the initial alignment portion):
#Finished.
# Total Time: 1429.47 seconds.
# Found 20000000 sequences.
# Sequences matched: 10173641
# Sequences not matched: 9826359
# Output written to /mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/crick/2_0.sam
validating PER (/mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/up/1.sam,/mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/dn/2.sam) mappings...
touch: cannot touch `/mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/up/1.sam': No such file or directory
touch: cannot touch `/mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/dn/2.sam': No such file or directory
cat: /mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/up/1.sam: No such file or directory
cat: /mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/dn/2.sam: No such file or directory
sort: open failed: /mnt/dznehomes/homes/ryand/Alignments/test_1/aligned/gnu/gen/watson/chr1/up/1.sam-temp1-87161: No such file or directory
A similar set of errors is issued for every chromosome and, indeed, those files don't exist (the directory skeleton is created, but never populated with files). I using a multifasta file as input, so might these be causing the behaviour or have I just incorrectly specified an option?
An unrelated question: Is there a way (as with direct use of gnumap) to simply save the hash to a file so it only needs to be created once. Recreating it for every sample that might need to be processed seems rather repetitive.
Thanks much
changjin
Dec 12, 2013, 3:07:23 PM12/12/13
to gnumap...@googlegroups.com, dpry...@gmail.com
The updated gnumap-bs pipeline has been fixed to resolve this issue. Now, it supports more general header in FASTQ format read.
Changjin Hong
Dec 12, 2013, 2:20:38 PM12/12/13
to gnumap...@googlegroups.com
The updated gnumap-bs pipeline has been fixed to resolve this issue. Now, it supports more general header in FASTQ format read. Thanks! Devon
--
---
You received this message because you are subscribed to the Google Groups "gnumap-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gnumap-users...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
Reply all
Reply to author
Forward
0 new messages