Using Illumina Masked References with BWA MEM
I have seen a couple articles where GATK recommends the usage of the Illumina masked reference genomes with either DRAGMAP or BWA MEM.
This article suggests that DRAGMAP now contains an option to run BWA MEM, and it's now compatible with Illumina masked graph based references.
When I downloaded DRAGMAP through conda, I don't see any option to use BWA MEM. Any guidance you can provide on how to get BWA MEM working with the Illumina masked references would be greatly appreciated. Thanks!
-
Hi pbrennan
I guess there is a misunderstanding on the BWA MEM fallback option. That option is not in DRAGMAP but it is in the wdl workflow that is used by the Terra workspace explained in the article. If you wish to use BWA MEM instead just use BWA MEM in your compute environment instead of DRAGMAP.
To use the same masked genome as DRAGMAP you have the following options present.
1- Use the GRCh38 reference genome with alt contigs present which is what we have in our resource bundle and mask it with the given bed file in the repository for DRAGEN. You may use bedtools or similar to do that.
2- Use GRCh38 reference genome that does not have any alt contigs. Additional masked regions can be checked from the above mentioned bed file.
The workflow mentioned in the article does not use any graph based references instead it uses the same linear reference genome. BWA MEM is not compatible with any graph based reference genomes.
I hope this helps.
Regards.
-
One last thing to add. Our team has tested BWA MEM against DRAGMAP using the workflow in the terra workspace and observed that the gap between 2 mappers gets closer but BWA MEM is still not functionally equivalent to DRAGMAP possibly due to differences between how each mapper handles mapping qualities and how DRAGMAP model is actually trained to assign mapping qualities to reads in certain regions of the genome.
I hope this helps.
Please sign in to leave a comment.
2 comments