package

1. pytorch
2. pytorch-geometric
3. tqdm
4. pandas
5. numpy
6. sklearn
7. matplotlib
8. seaborn
9. tensorboardX

run

1. python main.py
   1. random seed can be change inside main.py

complete pipeline from SC to graph to experiment

1. run preprocessing.py, it will analysis the SC and output pyg data
2. run data_split.py it will split the pyg data into train, valid(for fine tune), and test
3. run main.py