The scripts listed above are used to reproduce the results for TransOrGAN study. Due to the size limitation at GitHub, the data and model information are provided by request. To re-run these scripts, please make sure update the data and model path to your working directory.
- generate_train_set.py and generate_test_set.py are used for generating transcriptomic profiles for the training and test set using TransOrgan.
- control_cosine.py, mse.py and calculate_umap.py are used for calculating the cosine, root mean square error and UMAP projections for the training and test set.
- pathway_cosine.py and sex_specific36.py are used for calculating the cosine similarity for 87 CTD pathways and 36 sex related specific genes.