Todo:
- Change all files to load in the data in data/cohort_clin_data
- Create a makefile to automate the running of this code.
- Switch brca_regimens.rmd and process_drug_data.R away from using data lists, and use the individual datasets instead.
- It's not a rare event at all (>50 cases) for participants to have a report date which is after their death or loss to followup. Why is this? Does this indeed challenge our assumptions?
- Update all reports to read in the distinctly filtered survival datasets.
Curiousities:
- Are GENIE-MSK-P-0017706-T02-IM6 and GENIE-MSK-P-0017706-T01-IM6 totally identical?