tophat2 Reporting output tracks bam_merge failed

[Edden]

Hello,

I have a problem with tophat reproting output tracks. It might be similar to the one reported here. http://seqanswers.com/forums/showthread.php?t=24205.

However, in this thread the problem is solved with using tophat v2.0.6. In my case is is happening with exactly this version.

The error message is the following:
[2012-11-15 13:32:17] Reporting output tracks
        [FAILED]
Error running /ngsdata/ngs/tophat/tophat-2.0.6/tophat_reports --min-anchor 8 --splice-mismatches 0 --min-report-intron 50 --max-report-intron 10000 --min-isoform-fraction 0.15 --output-dir ./tophat_D6T90_R1_PE100_trimmed/ --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 10000 --min-segment-intron 50 --max-segment-intron 10000 --read-mismatches 2 --read-gap-length 2 --read-edit-dist 2 --read-realign-edit-dist 3 --max-insertion-length 3 --max-deletion-length 3 -z gzip -p24 --inner-dist-mean 50 --inner-dist-std-dev 20 --gtf-annotations /home/jlinde/Gff/Aspergillus_fumigatusa1163.CADRE.15_cufflinks.gff --gtf-juncs ./tophat_D6T90_R1_PE100_trimmed/tmp/Aspergillus_fumigatusa1163.juncs --no-closure-search --no-coverage-search --no-microexon-search --sam-header ./tophat_D6T90_R1_PE100_trimmed/tmp/Aspergillus_fumigatusa1163.CADRE_genome.bwt.samheader.sam --samtools=/usr/bin/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 /ngsdata/ngs/bowtie/indexes_bowtie2/Aspergillus_fumigatusa1163.CADRE.fa ./tophat_D6T90_R1_PE100_trimmed/junctions.bed ./tophat_D6T90_R1_PE100_trimmed/insertions.bed ./tophat_D6T90_R1_PE100_trimmed/deletions.bed ./tophat_D6T90_R1_PE100_trimmed/fusions.out ./tophat_D6T90_R1_PE100_trimmed/tmp/accepted_hits ./tophat_D6T90_R1_PE100_trimmed/tmp/left_kept_reads.m2g.bam,./tophat_D6T90_R1_PE100_trimmed/tmp/left_kept_reads.m2g_um.mapped.bam,./tophat_D6T90_R1_PE100_trimmed/tmp/left_kept_reads.m2g_um.candidates ./tophat_D6T90_R1_PE100_trimmed/tmp/left_kept_reads.bam ./tophat_D6T90_R1_PE100_trimmed/tmp/right_kept_reads.m2g.bam,./tophat_D6T90_R1_PE100_trimmed/tmp/right_kept_reads.m2g_um.mapped.bam,./tophat_D6T90_R1_PE100_trimmed/tmp/right_kept_reads.m2g_um.candidates ./tophat_D6T90_R1_PE100_trimmed/tmp/right_kept_reads.bam
Error: bam_merge failed to open BAM file ./tophat_D6T90_R1_PE100_trimmed/tmp/right_kept_reads.m2g_um.candidates1.bam

I've tried using --keep-tmp an the file bam_merge fails to open, does exist.

I am using a bunch of paired-end data, 100 bp, with ~20 Mio read pairs. Has been working fine for 28 out of 29 samples.

complete call looks like that:

tophat2 -G mygff .gff -g 1 --no-mixed --no-discordant -p 24 --b2-very-sensitive -o tophat_a myindex  a1.fq a.fq

[Edden]

Short update:

works perfectly with tophat1 /bowtie 1 and tophat2.04 /bowtie2.0.0-beta5

[Tara Ellison]

Edden, did you ever figure out the problem? I am getting the same error for all of my samples, 50bp single read (using tophat-2.0.6, and Bowtie 2.0.2 and Bowtie 0.12.8). I haven't tried tophat2.04 with bowtie2.00-beta5 yet - that's my next step.

[Edden]

Hi,

no final solution, but we figured out the number of parallel processes might be a problem (-p parameter). Tophat2 produces as many temporary bam files as processors your are using. It might happen that bammerge finally fails with too many files.
Unfortunately, I cannot give a maximal number of processors. In my case it worked with 12, but didn't work with 24. Think it depends on the size of the input (fastq) files.

Also using less processors is surprisingly faster in computation.

[Tara Ellison]

tophat confused by multiple fastq files

Hi, I got both tophat2 and earlier versions working after I merged my fastq files. It looks like the program was interpreting multiple files as PE files, and seeing errors when they weren't actually pairs. Also, I agree tophat worked better after I decreased the -p to 8.

[brooksma]

Thanks for this post!

I was trying to align 76base single reads using Tophat v2.0.7 and Bowtie 2.0.6 and was getting the BAM merge error. I could get it to work on my iMac with 3 cores bit not on my Linux server with 30 cores. I set the alignment on the linux server to 12 cores and it works, brilliant!