This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Cufflinks- found spliced alignment without XS attribute

I am getting the following command line message (various times) when running Cufflinks:

BAM record error: found spliced alignment without XS attribute

What does this mean? Is the program running correctly? Is the output reliable?

rna-seq cufflinks
• 6,586 views • link
updated by michael.ante michael.ante • written by Lindsay44 • Lindsay44

1 answer

Hi Lindsay44,

it means that your aligner did not report this optional attribute to the alignments. The alignments themselves can be correct and valid, but Cufflinks need this optional attribute. If you have a strand-specific library, you can circumvent this issue by providing Cufflinks the library-type parameter.

Otherwise, you'll need to re-run the alignment with a program which reports the XS attribute (e.g. Tophat2, STAR with the "--outSAMstrandField intronMotif" parameter, bbmap with the xstag setting...).

Cheers,

Michael

• 1 views • link
written by michael.ante michael.ante

Okay thank you.

Is there an option for this in the HiSat2 aligner?

How will my results be affected if I don't make this change?

• 1 views • link
written by Lindsay44 • Lindsay44

According to the HiSat2 Manual:

We can report this field for the canonical-splice site (GT/AG), but not for non-canonical splice sites. You can direct HISAT2 not to output such alignments (involving non-canonical splice sites) using "--pen-noncansplice 1000000".

I'd give STAR or BBMAP a try.

• 1 views • link
written by michael.ante michael.ante

It is a little late for me to switch at this point. If I don't change anything, will the results be inaccurate?

I have single end data- how should this be put into Hisat2?

• 1 views • link
written by Lindsay44 • Lindsay44

Log in to answer this question.

More posts like this