e-mail address: jiadong.yan@mail.mcgill.ca tensorflow 1.10 on GTX 1080 Ti mh_gagnn_spatial.py (1) inputs: all inputs are defined in function load_data() "train_path" is the path of the preprocessed brain data "label_path" is the path of the labels (2) outputs: "result.txt" file to record the training loss and testing loss and ten .mat files are the results of the modeled ten RSN spatial patterns mh_gagnn_temporal.py (1) inputs: all inputs are defined in function load_data() "train_path" is the path of the preprocessed brain data "label_path" is the path of the labels "spatial_p" is the modeled spatial patterns via spatial network which is also the input of the temporal network (2) outputs: "result.txt" file to record the training loss and testing loss and ten .mat files are the results of the modeled ten RSN temporal patterns