Vignette modification - seqlevelsStyle for non-hg19 bams

[biobenkj]

When calling variants from anything that doesn't have the default "chrM" in the BAM header, callMT will fail even if CHR = "MT" (or related) is specified (or the building of the index contains the appropriate MT contig) since seqlevels defaults to the "chr" annotations.

Current workaround is to specify:

seqlevelsStyle(MAlignments) <- "NCBI"

[ttriche]

shouldn't it be

seqlevelsStyle(MAlignments) <- "UCSC" 

if the goal is to get chrM in there?

[biobenkj]

During TallyVariants, it looks in the bam header and if it was aligned with anything that doesn't use the chrM annotation (e.g. b37), it can't find what it's looking for. The callMT function will work with the above workaround but the injectMtVariants will fail since it is now looking for chrM (as far as I can tell). What I'm currently doing is just reheadering the b37 aligned bams prior to import and all seems to be going well so far.

[ttriche]

crap, this needs to be fixed. maybe when we merge in the false positive filters from Xiaowu we can hit this and T/N bulk at the same time. aargh

…

--t

On Mon, Sep 17, 2018 at 12:30 PM Ben Johnson ***@***.***> wrote: During TallyVariants, it looks in the bam header and if it was aligned with anything that doesn't use the chrM annotation (e.g. b37), it can't find what it's looking for. The callMT function will work with the above workaround but the injectMtVariants will fail since it is now looking for chrM (as far as I can tell). What I'm currently doing is just reheadering the b37 aligned bams prior to import and all seems to be going well so far. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#13 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAARIrlJW38Z4Oe_FJQQXFncqS3lM7pOks5ub847gaJpZM4Wp0ZO> .