Code for the reproducibility of MHC-I genotype restricts the oncogenic mutational landscape.

Requirements

All figures are generated in either Python or R. The following libraries are required:

• Python
  • pandas
  • scipy
  • matplotlib
  • Bio
  • sklearn
  • seaborn
• R
  • parallel
  • lme4
  • mgcv
  • xtable

Furthermore, all figure generating scripts are based on pre-processed data. Please contact me to acquire access to the data.

• Once data is acquired, create update the imports/import.pynb notebook with the location of the data.

Data processing

Example code for processing the raw data is included in the data_processing directory. Thousands of CPU hours are required to construct the pre-processed data files required to generate the figures. If you want to re-generate the processed data, this code will be a good guideline but pipelines will have to be set up for your own cluster architecture.

Also, the R-based statistical models take some time to run. I have included an independent R script to generate the figures from the model in addition to the jupyter notebooks.