trouble with bwakit - where are the options?

[says_anova]

Hi all,

I'm wondering if anyone has some experience with bwakit v0.7.12 and can chime in.

The command run-bwamem within bwakit seems to have many fewer options than previous stand alone versions of bwa. Most notably, I would like to pass the option -M to bwamem, which marks shorter split hits as secondary. However, when I execute run-bwamem with this option, it throws an error:

/r/bwa-kit/run-bwamem -o test -t8 -MHR"@RG\tID:run1\tSM:${sample}\tLB:LaneX\tPL:Illumina\tPU:none" /r/bwa-kit/hs38DH.fa LP6005206-DNA_A06.1.fastq LP6005206-DNA_A06.2.fastq | sh

Stout: Unknown option: M

Is it the case that this option simply doesn't exist for the version of BWA mem that bwakit calls?

[N311V]

I've used bwa mem but not via that kit.

Does running /r/bwa-kit/run-bwamem give you the available options?

[says_anova]

Yes, it does list some options, but they are far more limited than what one sees running bwa mem. I suppose that's a big clue, but I assumed those options existed on top of what you can already pass to bwa mem. Seems like that's not the case, though. A big minus for bwa-kit.

Can anyone else confirm or deny?

Thanks all!

[GenoMax]

If you look at the run-bwamem script you can see the options it accepts (which is a sub-set of bwa mem): https://github.com/lh3/bwa/blob/mast...kit/run-bwamem

Looks like Heng Li developed this as an end-to-end solution for HLA typing analysis (https://github.com/lh3/bwa/blob/master/bwakit/README.md).

[blancha]

@Genomax is correct.
bwakit is nothing more than a collection of Perl scripts written to process data with bwa in the context of a specific project.
It seems simpler and more direct to use bwa mem directly, without using the bwakit scripts, especially if your project is different from theirs.
Nevertheless, the Perl scripts can easily be modified, if you want to continue working with bwakit for your own project.

I've forked the bwakit GitHub repository, and added the -M option.

https://github.com/blancha/bwa/tree/master/bwakit

The usual caveat applies that my modifications may always introduce a new bug, but that any bug detected can easily be fixed.

[says_anova]

Thanks blancha, Genomax, for the clarification. Sounds like the simplest thing for us would be to use the stand alone BWA software for the time being.