This repository contains all jupyter notebooks to reproduce the results of the single-cell data analysis from:
M. Thomas, R. Brabenec et al., "The role of single cell transcriptomics for efficacy and toxicity profiling of chimeric antigen receptor (CAR) T cell therapies", Computers in Biology and Medicine (2025).
The notebooks contain code for the following analyses:
- Clinical efficacy and safety from clinical patient outcome data (Fig. 1)
- CAR target expression across tumor tissues (Fig. 2 and Suppl. Fig. 1)
- CAR target expression across healthy tissues (Fig. 3)
- Correlation of CAR target expression profiles with observed clinical toxicity effects (Suppl. Fig. 2)
Note that the analysis was done in python 3 using scanpy v.1.4.6 to 1.6.1 and anndata v.0.7.1 to 0.7.5 unless otherwise stated. Some functions may have changed when using other package versions.
Additionally, numpy (v. 1.18.2+), pandas (v. 1.3.5+) and scipy (v. 1.4.1+) are required. Numeric results can vary depending on package versions and e.g. minor results. All figures were plotted using matplotlib and seaborn.
If the materials in this repository are of use to you, please consider citing the above publication.