TxGNN
TxGNN copied to clipboard
mims-harvard
TxGNN: Zero-shot prediction of therapeutic use with geometric deep learning and clinician centered design
When I am trying to load the pretrained graph then I get following error. TxGNN.load_pretrained('./TxGNNExplorer') dgl._ffi.base.DGLError: [14:31:29] /tmp/dgl_src/src/graph/heterograph.cc:120: Check failed: !rel_graphs.empty(): Empty heterograph is not allowed. The same error appears...
I've been looking into the code, and this [line](https://github.com/mims-harvard/TxGNN/blob/1000aac7120e0022af09a3ef93dba219e16a094b/txgnn/utils.py#L755) seems a bit puzzling (numerical experiments don't show any difference in the model's performance w/ and w/o this line). What is...
I'm trying to install given the instructions. I make it as far as ``` conda create --name txgnn_env python=3.8 conda activate txgnn_env # Install PyTorch via https://pytorch.org/ with your CUDA...
Hi, When I try to run the cardiovascular split I encounter the following problems: - The object `TxData` downloads `node.csv` and not `nodes.csv` https://github.com/mims-harvard/TxGNN/blob/1000aac7120e0022af09a3ef93dba219e16a094b/txgnn/TxData.py#L25 - When the object `DataSplitter` calls...
Why in line 679 we're unpacking 3 values: ` _, scores, _ = self.retrieve_gates_scores_penalties()` while the retrieve_gates_scores_penalties function only return 2 items in line 676 ` return whole_graph, test_graph` P.S:...
Not sure how to do this, since there is not function on TxEval only in `utils.py`. Can I somehow call the [GNNExlainer from DGL](https://docs.dgl.ai/en/1.1.x/generated/dgl.nn.pytorch.explain.GNNExplainer.html) directly on the model? :thinking:
This PR provides an out-of-the-box the implementation of a RESTful API that exposes TxGNN's functionality for zero-shot therapeutic predictions and explanations using geometric deep learning (GNN). This API is deployed...
Hi, I can't find the dgl 0.5.2 for installation, could you please provide some more info in this regard? Best Regards Anahita
Metadata
Owner
Metadata
TxGNN: Zero-shot prediction of therapeutic use with geometric deep learning and clinician centered design