/MTseeker

mitochondrial variant analysis tools

Primary LanguageR

MTseeker

MTseeker is a toolkit for mitochondrial variant analysis in the Bioconductor environment. Given suitably aligned reads, it will call variants (and/or mitochondrial copy number), plot them for multiple samples, and provide basic functional annotation for coding variants.

If we were cool people, it would have a mitochondrion stuffed into a hexagon, but we aren't, so it doesn't... yet.

Build Status

How to install

devtools::install_github("trichelab/MTseeker")
# or:
install.packages("BiocManager")
BiocManager::install("trichelab/MTseeker")

How it works

Visit the vignettes for more details, but suppose we have a bunch of renal oncocytomas (known to be driven early in development by mitochondrial mutations) and matched adjacent normal kidney samples (let's call them "RO" and "NKS").

MTreads <- getMT(BAMfiles)

The first and simplest thing we might like to do is get mitochondrial reads from exome sequencing BAMs (or, really, any BAM -- it doesn't have to be any particular sequencing technology, though at present we support human mitogenomes far better than any other organism). Above, the getMT function does that.

Now we might also like to estimate from the mt/nuclear read support what the rough ratio of mitochondria in the tumor tissue is compared to the normal tissue. Since we track the read support for each upon load, we can use that:

mVn <- Summary(MTreads)$mitoVsNuclear
names(mVn) <- names(MTreads) 
CN <- mVn[seq(2,22,2)]/mVn[seq(1,21,2)] 
mtCN <- data.frame(subject=names(CN), CN=CN)

library(ggplot2) 
library(ggthemes)
p <- ggplot(mtCN, aes(x=subject, y=CN, fill=subject)) + 
  geom_col() + theme_tufte() + ylim(0,5) + 
  ylab("Tumor/normal mitochondrial ratio") + 
  ggtitle("Mitochondrial retention in oncocytomas")
# ggsave("inst/extdata/mtCN.png") 
print(p) 

mitochondrial CN

Now we probably want to call variants, and then plot them (filtered):

variants <- callMT(MTreads)
plot(filt(variants))

That looks like so:

mitochondrial variants for 22 samples

It would be nice to have some idea what this means for each patient:

mitochondrial variant fallout in patient 1

More to come! But this is the basic idea.