Tool or R package for computing Tumor Mutational Burden?

Question

[13636443958]

Tumor Mutational Burden is the 'number of variants per megabase' but, how can I compute this? Are there tools, software or R packages that can obtain this information from the vcf files? I have no idea about this!

Answer 1

[Kevin Blighe]

You can code this - there does not have to be a package. There was also a recent previous thread about this, if you can please do a quick search?

Essentially, there is no single way that people are calculating TMB. Some people only look at somatic missense mutations, while others look at missense + nonsense.

From my own perspective, even intergenic and intronic somatic mutations should be included as these are reflective of unstable genome maintenance processes. However, one can quite easily just look at missense and nonsense and then correlate these to, e.g., TILs, neoantigens, etc (as these will be more likely to be antigenic).

Kevin

[atheeth14]

Hi Kevin, I intend to calculate TMB per Mb of the sequence I have. I'm working with R and the aligned BAM files. My understanding is that to write a code for the same, wouldn't I require a set of given mutational signatures (the non-synonyms mutations in this case)? Could you elaborate on how one would go about coding a package to measure TMB in terms of per Mb?

Thank you, Atheeth

[Kevin Blighe]

Hi Atheeth, you could do this via a for loop in R that iterates over each megabase region and counts the mutations in each. Using GenomicRanges may make this easier; however, if you are new to GenomicRanges, the learning time may add significant extra time to the work.