extended cigar in bwa-mem

[dovah]

Hi all,

I am aligning pacbio reads to a reference genome using bwa mem. Do you know how to output an extended CIGAR in output sam file from bwa v0.7.13?

My command is:

bwa mem -x pacbio refgenome.fasta reads.fastq > output.sam

Thanks in advance

[dpryan]

What exactly do you mean by "extended CIGAR"? bwa mem will output SAM files with an appropriate CIGAR string already.

[Brian Bushnell]

He probably means cigar strings with X and = symbols instead of M, which are very handy.

[dovah]

Yes, this is what I mean by "extended cigar". By the way, I found a tool that can do that a posteriori, in case someone else is interested: SamFixCigar (http://github.com/lindenb/jvarkit )

[sklages]

Are there any plans to extend bwa to write X/= instead of M? At least optional?

[Brian Bushnell]

You can now use BBMap's reformat.sh to add those:

reformat.sh in=mapped.sam out=extended.sam sam=1.4

I have not tested it extensively on things like hard-clipping but it should generally work.

[sklages]

Brian,

thanks for the suggestion. I will try :-)

Nevertheless I'd wish bwa would provide an option to either use M or X=.
Just to avoid more I/O due to another conversion ...

Sven

[sklages]

After studying some docs I do see that with optional MD:Z I am even as flexible as with X=, with the advantage of not having too "complicated" CIGARs. :-)