A Highly Trustworthy Multimodal Learning Method on Omics
To get start with MoTIF, please follow the instructions below.
pip install -r requirements.txt
git clone git@github.com:YuxingLu613/MoTIF.git
The BRCA and ROSMAP dataset can be found in MOGONET. All the preprocessed datasets can be found in MoTIF Datasets. The pan-cancer dataset can be fetched in TCGA program .
To train the model from stratch with defult settings, you can run the code
python main.py
To train the model on a specific dataset, you can add argument
python main.py -task ("BRCA"|"ROSMAP"|"KIPAN"|"LGG"|"CCM")
If you just want to test the saved model in checkpoints/, you can add argument
python main.py -test_only True
If you want to change the number of input modalities, you can change the "uni_modality", "dual_modality", "triple_modality" argument
python main.py -uni_modality True -dual_modality False triple_modality False
If you want to change the hyper-parameters in the model, you can edit the train/train.py.
TBA
If you have any questions, please feel free to get touch with me, my email is yxlu0613 AT gmail DOT com