SIRT4 comutations in CRC

Joshua Cook
6/15/2020

Directive from Prof. Kevin Haigis

A very important collaborator is trying to finish up a grant on SIRT4
in colon cancer. SIRT4 mutation is rare, but I noticed from a quick
check of cBioportal that when SIRT4 is mutant, KRAS mutations are
infrequent (20%), PIK3CA mutations are frequent (almost 50%), and BRAF
mutations are frequent (26%).

1. Could you please do a more rigorous analysis of this in your
   larger dataset? I believe that PI3K pathway mutations are even
   more frequent than just looking at PIK3CA. Could you include PTEN
   and PIK3CG in a separate analysis?
2. We would like to include SIRT4 deletion in addition to mutation,
   since this is a more frequent event. Is that possible?
3. Also, my suspicion is that BRAF is enriched because it co-mutated
   with PIK3CA. Would you agree?

Analyses

Follow the steps listed below to re-run the analysis on O2:

1. Run Rscript src/get-data-from-comutation-project.R to get the data
   from the KRAS comutation analysis.
2. Run Rscript -e rmarkdown::render('src/sirt4-comutation-analysis.Rmd')

Conclusions

There were patterns of comutation between SIRT4 and the key members of
the PI3K signaling pathway, particularly with PIK3CA and PIK3CG.

These genes were also identified as strong predictors of a SIRT4
mutation in a binomial model.