xomx is an open-source python library providing data processing and machine learning tools for computational omics, with a particular emphasis on explainability.

It relies on AnnData objects, which makes it fully compatible with Scanpy.

xomx is currently in beta version.

---

Install

Option 1: pip

pip install git+https://github.com/perrin-isir/xomx

Option 2: conda

git clone https://github.com/perrin-isir/xomx.git
cd xomx

Choose a conda environmnent name, for instance xomxv.
The following command creates the xomxv environment with the requirements listed in environment.yaml:

conda env create --name xomxv --file environment.yaml

If you prefer to update an existing environment (existing_env):

conda env update --name existing_env --file environment.yml

To activate the xomxv environment:

conda activate xomxv

Finally, to install the xomx library in the activated virtual environment:

pip install -e .

---

Tutorials

Tutorials (in xomx/tutorials/) are the best way to learn to use xomx.

List of tutorials:

• xomx_kidney_classif.md, and a shorter notebook version: xomx_kidney_classif.ipynb
  goal: use a recursive feature elimination method on RNA-seq data to identify gene biomarkers for the differential diagnosis of three types of kidney cancer

• xomx_pbmc.md
  goal: follow the single cell RNA-seq Scanpy tutorial on 3k PBMCs, except for the computation of biomarkers for which recursive feature elimination is used

---

Citing the project

To cite this repository in publications:

@misc{xomx,
  author = {Perrin-Gilbert, Nicolas and Vibert, Julien and Vandenbogaert, Mathias and Waterfall, Joshua J.},
  title = {xomx},
  year = {2021},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/perrin-isir/xomx}},
}