Integration and analysis of Lung macrophage cells from patients with COVID-19, Idiopathic pulmonary fibrosis, COPD.
The following notebooks described the data integration, data exploration and proximity analysis steps of Methods as implemented in
Wendisch, Dietrich, Mari, Stillfried et al. (2021). (in press)
Jupyter notebooks located in the directory notebooks
describe steps for data of external samples and BAL (this study).
Pre-processed datasets and the integrated embedding and zipped and available in h5ad format in Dropbox (~14.3 GB).
To skip pre-processing and integration, please update the data
directory with those.
Main processing steps are:
- Data Preparation of Adams/Morse/Bharat and our work (filtering, normalization). H5AD files are generated for each dataset using the public raw data, and then deployed in
data
. - notebook Data Integration of all datasets using (scVI), and patient/sample IDs as batches.
- notebook UMAP visualization and gene expression analyses using macrophage gene groups post-processing.
- notebook Counting and quantification of link between macrophages from COVID-19 patients and IPF, using distances in the kNN graph from the integrated cells as a reference for mapping conditions (i.e. proximity analysis).
To install environment and relevant dependencies, please clone and execute the following command.
conda env create -f environment.yml
Option 1) Install jupyter
in the same environment.
conda activate covid_macrophages_integration
conda install -c conda-forge jupyterlab
Option 2) Add kernel to jupyter running in another environment.
conda activate covid_macrophages_integration
conda install -c anaconda ipykernel
python -m ipykernel install --user --name=covid_macrophages_integration
Troubleshooting: Please open an issue.
License: MIT.