Tophat failure: Reporting output tracks and bam_merge

[Jerry_Zhao]

Something wrong with bam_merge?

TopHat run (v2.0.6)
Bowtie version: 0.12.7.0
Samtools version: 0.1.18.0


tophat --bowtie1 -o ./SRR486241 --solexa-quals -p 10 -g 1 --no-coverage-search --no-novel-juncs --library-type fr-firststrand -G genes.gtf GRCh37 ESC_1.fastq ESC_2.fastq





Reporting output tracks
[FAILED]
Error running /usr/local/bin/tophat_reports --min-anchor 8 --splice-mismatches 0 --min-report-intron 50 --max-report-intron 500000 --min-isoform-fraction 0.15 --output-dir ./SRR486241/ --max-multihits 1 --max-seg-multihits 10 --segment-length 25 --segment-mismatches 2 --min-closure-exon 100 --min-closure-intron 50 --max-closure-intron 5000 --min-coverage-intron 50 --max-coverage-intron 20000 --min-segment-intron 50 --max-segment-intron 500000 --read-mismatches 2 --read-gap-length 2 --read-edit-dist 2 --read-realign-edit-dist 3 --max-insertion-length 3 --max-deletion-length 3 --bowtie1 -z gzip -p10 --inner-dist-mean 50 --inner-dist-std-dev 20 --gtf-annotations genes.gtf --gtf-juncs ./SRR486241/tmp/genes.juncs --no-closure-search --no-coverage-search --no-microexon-search --solexa-quals --library-type fr-firststrand --sam-header ./SRR486241/tmp/GRCh37_genome.bwt.samheader.sam --report-discordant-pair-alignments --report-mixed-alignments --samtools=/Users/yingtao/Jerry/Tools/samtools/samtools --bowtie2-max-penalty 6 --bowtie2-min-penalty 2 --bowtie2-penalty-for-N 1 --bowtie2-read-gap-open 5 --bowtie2-read-gap-cont 3 --bowtie2-ref-gap-open 5 --bowtie2-ref-gap-cont 3 GRCh37.fa ./SRR486241/junctions.bed ./SRR486241/insertions.bed ./SRR486241/deletions.bed ./SRR486241/fusions.out ./SRR486241/tmp/accepted_hits ./SRR486241/tmp/left_kept_reads.m2g.bam,./SRR486241/tmp/left_kept_reads.m2g_um.mapped.bam,./SRR486241/tmp/left_kept_reads.m2g_um.candidates ./SRR486241/tmp/left_kept_reads.bam ./SRR486241/tmp/right_kept_reads.m2g.bam,./SRR486241/tmp/right_kept_reads.m2g_um.mapped.bam,./SRR486241/tmp/right_kept_reads.m2g_um.candidates ./SRR486241/tmp/right_kept_reads.bam

Error: bam_merge failed to open BAM file ./SRR486241/tmp/right_kept_reads.m2g_um.candidates3.bam


The input files are very large. There are about 200 million pairs of paired-end reads.
When I tried the same parameter using the first 10,000 pairs, everything was fine and there was no failure.
Therefore, is it because the input FASTQ file is too large?

[westerman]

>Therefore, is it because the input FASTQ file is too large?

Well, that would be the very root cause, yes. Smaller input equals less load on the system overall. Now as if bam_merge itself is having problems ... that is a different issue. It could be simply that you are running out of disk space, disk quota, an network file error, memory constraints, etc instead of the bam_merge program per se. Are there any other error messages preceeding the one that you posted?

[Jerry_Zhao]

We have 2T hard drive and the input files are 67G + 67G , therefore the disk space is unlikely the problem.
We have 32G memory and the tophat only use half of them.

I have used the same parameter for several datasets with 20 million single-end reads last week, and everything was fine.


I have changed the parameter -p 10 to -p 7, and run tophat again.
I will update the result today or tomorrow.

[labrat73]

any news?

Jerry,

Having similar problem with TopHat 2.0.6. Have 2T hard drive and 96G memory. Input files are 20G + 20G. I too, have used the same parameter (-p 25) for smaller datasets with no problem. Any results when you decreased the number of threads from 10 to 7?

Appreciate any information...

[Jerry_Zhao]

Hi Labrat73,

Sorry that I forgot to update.

I only changed the parameter -p 10 to -p 7, and no error message occurred.
Nevertheless, I do not have high confidence about the results.

Therefore, I am going to try STAR, another RNA-seq aligner.
I will update after I get the STAR results and compare with the two.

Best,
Jerry

[Rui]

Hi Labrat73,

Sorry that I forgot to update.

I only changed the parameter -p 10 to -p 7, and no error message occurred.
Nevertheless, I do not have high confidence about the results.

Therefore, I am going to try STAR, another RNA-seq aligner.
I will update after I get the STAR results and compare with the two.

Best,
Jerry

Hi,

I am getting into that error for one of my samples. So the trick is to lower the number of threads to solve it. Did I read it correctly?

[Jerry_Zhao]

Yes.
Reduce the number of threads to no more than 10, there will be no error message during running.

Nevertheless, no error message does not mean that the mapping is correct. This is my opinion.

[Rui]

Jup, the job went through without crash.