This repository contains all scripts used in the preprint "The shared neoantigen landscape of MSI cancers reflects immunoediting during tumor evolution". This includes data anonymization, processing and generation of figures.

To re-generate data and plots, run run.sh from this directory.

To run the script for rerunning our analysis on processed data, first perform the following steps. In addition to standard GNU/Linux tools the following are necessary for use.

Set up a conda environment:

conda env create --name neomsi --file=env.yml conda activate neomsi

Install NetMHCpan (instructions: http://www.cbs.dtu.dk/services/doc/netMHCpan-4.0.readme), and add netMHCpan to your path:

export PATH=/path/to/netmhcpan/bin/:$PATH
export NETMHCpan=/path/to/netmhcpan/<Linux/Darwin>/

once installation is done, you can run run.sh to rerun analysis and figure generation.

Change directory to ReFrame. To run ReFrame on example data, run the following:

r -f reframe.R

This will process all .xlsx files containing raw fragment length data in the in directory and deposit processed allele frequency data to the out folder. Example input and output files are provided in the respective folders.