Deconomix-case-study

An example case study for the Deconomix python package. It demonstrates the capabilities of the package in a breast cancer scenario, including data curation, preprocessing, processing and visualizations.

Prerequisites

While the analysis part in this study is computable on standard personal computer, the preprocessing of the data is not feasible on such a system. Therefore we strongly advise you to run this study on a cpu cluster with around 100GB of RAM, which will also drastically reduce the compute time for the hyperparameter search, where a multitude of models have to be trained. We provided an example Dockerfile as well.

Install instructions

To run this case study on your own system, you can clone this repository to your disk. The source of the python package is included as a git submodule, therefore you should run following commands to run the analysis:

git clone https://github.com/Deconomix/Deconomix-case-study.git
cd Deconomix-case-study
git submodule update --init --recursive

Then you can create a Python virtual environment or a Docker container and run this command inside, which will then install Deconomix from the git submodule and all other requirements needed for data curation and preprocessing.

pip install -r requirements.txt

Dataset

In this study two public datasets will be used that have to be downloaded separately:

• DISCO Breast Cancer/Healthy Single-Cell Expression Data: Download disco_breast_v01.h5ad from https://zenodo.org/records/7396984. Move this file to the ./Data/DISCO directory.
• TCGA-BRCA Breast Cancer Bulk Expression Data. Download via Download_TCGA.py which implements an API call to the GDC database. It will save the dataset to ./Data/TCGA automatically, if you execute the script in the root of this project.