This repository contains reproducible analyses for "Recurrent repeat expansions in human cancer genomes." Read the paper on Nature.com.
The ehdn-lrdn-tool/ directory links to the ExpansionHunter denovo local read depth normalization tool. Instructions for using the tool are within the submodule.
The ehdn-lrdn-benchmark/ directory contains a Snakemake workflow for benchmarking the performance of the local read depth normalization method.
The analyses can be found in the analysis/ directory. To run the analyses, first install the required packages using conda:
$ conda env create -f environment.yml
Then, activate the conda environment and add it as a Jupyter kernel:
$ conda activate cancer-rre
$ python -m ipykernel install --user --name=cancer-rre
Finally, start a Jupyter notebooks instance to run the analyses interactively:
$ jupyter notebooks
Make sure to select cancer-rre as the kernel.