Code for paper: Multimodal framework for the joint analysis of single-cell RNA and T cell receptor sequencing data predicts T cell response to cancer immunotherapy
The code has been developed and tested on a high-performance computing system running Ubuntu 20.04.4 LTS. The system is equipped with an AMD EPYC 7513 32-Core Processor, 2 TB of RAM, and eight NVIDIA RTX A6000 GPUs (each with 48 GB VRAM), using CUDA version 12.2 and NVIDIA driver version 535.104.12. The demo application of our pipeline can be run on significantly less powerful hardware, requiring:
- Linux-based system
- at least 20 GB of free disk space
- one GPU
- an internet connection
Follow the two steps illustrated below
- create a conda environment using
environment.yaml(all dependencies are included; whole process takes about 5 min):
conda env create -f environment.yml
- install the current package in editable mode inside the conda environment:
pip install -e .
To run the models, please follow these steps:
- Preprocess the RNA-seq and TCR-seq data (see example:
./analysis/HNSCC/data_preprocess/data_processing.py) - Update the paths in ./run.sh
- Execute the script to start the run
./run.sh
Illustraive figures: made using powerpoint
Pointers for nonillustrative figures:
./analysis/HNSCC/2.0.eval_data_explore.py: Fig.3 (a, c, d, e, f, g), Supp Fig.4./analysis/HNSCC/2.1.eval_rna_pairwise_dist.py: Fig.3b, Supp Fig.2a./analysis/HNSCC/2.2.eval_gen.py: Fig.4, Fig.5a, Supp Fig.5, Supp Fig.7, Supp Fig.8./analysis/saliency: Fig.5 (b,c,d), Supp Fig.9./analysis/pan-cancer/2.1.evaluation.py: Supp Fig.3c, Supp Fig.6./analysis/pan-cancer/2.0.eval_rna_pairwise_dist.py: Supp Fig.3b