bwa.kit: run-bwamem or bwa mem

Question

[iraun]

Hi there,

I have recently realised about bwa.kit toolkit. BWA.KIT

My questions are:

• What is the difference between running bwa mem and running run-bwamem ? As far as I have read in the manual, run-bwamem performs bwa mem mapping and then a post-processing step. Is the mapping step in run-bwamem equal to bwa mem ? Also I'm aware of the run-bwamem options for the mapping step like marking duplicates, removing adapters and so on, but I'm wondering if there is any difference in the mapping algorithm.
• Is there any case when post-processing step is NOT recommended?
• Is there any case when bwa mem should be used instead of run-bwamem ?

Thank you in advance,

[abascalfederico]

Hi, what I understand is that run-bwamem runs not only bwa mem but also a pipeline that handles the multiple alternative contigs present in build 38 of the human genome. When there are alternative contigs for a given genomic region, you may get that a uniquely mapping read is mapped to different homologous alternative contigs. Without run-bwamem such a read would be considered non uniquely mapping. The pipeline also deconvolutes the mapping of reads to multiple HLA haplotypes to distill information on which are the two most likely HLA haplotypes. There may be other features within run-bwamem I am not aware of.