This is the code for the L45 project "Incorporating biological pathways information to aid deep learning on scRNA-set data".
To run the experiments in the paper, execute the benchmarking.py script.
python3 benchmarking.py
Plots can then be created using the plot_results.py script with the name of the csv file created by the benchmarking.py script.
python3 scripts/plot_results.py experiments/all_models.csv
There is no command line interface, so to change the experiment setup, you need to edit the ExperimentConfig object in the benchmarking.py script directly. For example, to change the output csv location, change the ExperimentConfig.name parameter in the instantiation.