Cufflinks error: BAM record error: found spliced alignment without XS attribute

Question

[Rashedul Islam]

I got RNA-seq bam files that are aligned by BWA. When I am using these bam files in cuflinks I got following errors.

"BAM record error: found spliced alignment without XS attribute"

My command was:

./cufflinks -p 24 -G /path/Homo_sapiens.GRCh37.69.gtf -o /path/Cufflink_Out/ /path/test.bam

Can anyone help me?

Answer 1

[Manvendra Singh]

Was your data stranded?

If yes then you should specify the library type when you run cufflinks

BWA does not store the strand tag, but TopHat and STAR do.

It's better if you map with TopHat or STAR,

With bwa, please specify the strand while running cufflinks.

[Rashedul Islam]

Thanks a lot Manu. I did not map the reads to the reference genome and got the bam files from other source. So, I do not know whether the data is stranded or not. Right now it is not possible for me to remap with TopHat or STAR. However, I tried with specifying 3 different library types to run cufflinks e.g., fr-unstranded, fr-firststrand and fr-secondstrand. Only fr-secondstrand gave me the FPKM values of gene, isoforms and transcripts. fr-unstranded gave error BAM record error: found spliced alignment without XS attribute and fr-firststrand was aborted and the last line was 7ffcbb0b3000-7ffcbb0b4000 rw-p 00000000 00:00 0 Aborted.

So, can I rely on the FPKM values of fr-secondstrand library type? Please let me know. Thanks in advance.

[Manvendra Singh]

Dear Rashedul,

I am not sure about it. But I think Istvan Albert, Pierre Lindenbaum or Devon Ryan could probably answer this

[Manvendra Singh]

If possible get fastq files from source and then map with tophat or STAR

[Ron]

Hi Manu,

I am getting this similar thing.Is this an error because I have got my output files generated?

Also I have used STAR as my aligner and still getting this issue.Let me know your thoughts.

[Rashedul Islam]

I solved the error by adding the library type in Cufflinks. My command was:

./cufflinks -p 12 --library-type fr-firststrand -G /path/Homo_sapiens.GRCh.gtf -o /path/ressult/ /path/file.bam

[Ron]

DId you find the change in results?I saw that output files are still generated in the case where it shows an error.

[Rashedul Islam]

I did not get any error finally.