tophat2/samtools

[ajgentles]

We have been trying to move from tophat1 to tophat2 utilizing a transcriptome as well as genome reference, and are having some performance and output issues. RNA-seq data that previously took about 12 hours to map now takes >4 days (stopped the runs at that point), with tophat_reports apparently being the culprit. Providing just a few hundred reads still takes much longer than expected and ends with an error in tophat_reports. The reported error is typically something like

Warning: mapped sequence without CIGAR (DJG84KN1:201:C0CKNACXX:7:1101:1815:2215)
Warning: mapped sequence without CIGAR (DJG84KN1:201:C0CKNACXX:7:1101:1815:2215)
Warning: mapped sequence without CIGAR (DJG84KN1:201:C0CKNACXX:7:1101:2101:2179)
Warning: mapped sequence without CIGAR (DJG84KN1:201:C0CKNACXX:7:1101:2101:2179)
Warning: mapped sequence without CIGAR (DJG84KN1:201:C0CKNACXX:7:1101:2307:2249)
Warning: mapped sequence without CIGAR (DJG84KN1:201:C0CKNACXX:7:1101:2307:2249)
Error: CIGAR and sequence length are inconsistent!(TGCTCTTCCGATCTGCCCCCTTAAACACCATTTTCCCTCCAGGACCACCTTGGTTTCTAGGCACTGTGGTTCTTGGCAGGGGCTGTCTTAGG)

This looks like it's possibly an incompatibility with samtools but we have tried versions 0.11 through 0.18 with no change. I wondered if anyone can help narrow down the issue ?

[kristofit]

Hi,

I got the same problem using tophat 2.0.6. (and bowtie 2.0.2) and can not find a way to solve the problem.
- error log is:
Error running /PATH/tophat.dir/bin/tophat_reports
Error: CIGAR and sequence length are inconsistent!(TTCAAACAAAATCGAATCCTGAAAGAGTAGAAGGGGAGCGGTGAGAGGAGGAGGAGGAGGAAGAGGAGGAGGGGGGCAGTCCTCCCCGAGCTAAAAACCTC)
Somebody did solve this problem ?

[tschauer]

Hi,

Same problem...

Did you guys solve it?

[ajgentles]

I found that if you tell tophat2 to generate a
transcriptome index by supplying a GTF file, but the index already
exists, it bombs out. You have to make sure not to tell it the GTF file
again.

We've basically abandoned tophat2 in favour of STAR these days.

[tschauer]

thanks

samtools was not correctly installed

without GTF it works

[joseph.troy]

Question about removing the GTF file

Thanks all for the information! I'm having the same problem. I tried removing the GTF file, but perhaps did it wrong (see with and without below). If possible can you share your command lines without the GTF file?

My command with GTF file...
tophat -p 8 -o tophat.out --library-type fr-firststrand -G ~/jmt/projects/run3_01/mm9gtf/genes.gtf --transcriptome-index ~/jmt/projects/run3_01/mm9transcripts/transcriptome ~/jmt/projects/run3_01/mm9genome/genome test.fastq


My command without the GTF file...
tophat -p 8 -o tophat.out --library-type fr-firststrand --transcriptome-index ~/jmt/projects/run3_01/mm9transcripts/transcriptome ~/jmt/projects/run3_01/mm9genome/genome test.fastq

-Thank you!!

[pengchy]

Hi all,

this problem seems still unresolved.
GTF was recommended by Tophat2 paper if available.
In my case, only one line has the problem in one bam file:

        73      scaffold210     158009  50      30M26178N70M    *       0       0       GTACGAGTCGTTCTGCCGGCCGCCGTGCTCGGAGTCGCCGTTGACGATCCAGACGATGTGCGGCGCGGGCTTGGCGGAGCCGGAGCTGCAGTTGGCGTGC  $#$%&&&#%$"&"%("')&!&)''&!"%%(""!$!"!!"""#%%"$%%%%$"!"!!!!!"%!"%$"#%%%%""%%!!!!"$%"!!"%$$!!!!!!!!!!!  AS:i:-12        XM:i:2  XO:i:0  XG:i:0  MD:Z:78C18C2    NM:i:2  XS:A:-  NH:i:1

Where, the read name was not printed successfully.

I filtered this line by:
awk '$1!~/^[0-9]/'