Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clintox dataset example #271

Closed
wants to merge 5 commits into from
Closed

Conversation

kshitij-v-mehta
Copy link
Collaborator

No description provided.

Kshitij V. Mehta added 2 commits July 26, 2024 14:07
Several molecules seem to have invalid an SMILES string. Added a
try..catch to ignore those molecules.
@kshitij-v-mehta
Copy link
Collaborator Author

kshitij-v-mehta commented Jul 26, 2024

I formatted using black, but CI still fails. Any suggestions?

]
},
"Variables_of_interest": {
"input_node_features": [
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Running the code in debug mode, may you check what is the shape of the tensor data.x? For this problem, data.x should have only one column.
If this is the case, then we should use

"input_node_features": [
  0
  ],

0
],
"type": [
"graph"
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it is correct to use only one head to make graph-level predictions.
However, it seems like the predictions require performing a binary classification problem.
Currently, HydraGNN is set up to perform only regression model.
To adapt the code for classification, changes are minor, but need to be made. Namely:

  • use sigmoid as activation function at the last layer
  • use binary cross entropy as loss function for training, validation, and testing

Copy link
Collaborator

@allaffa allaffa left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

A few comments. Let's discuss it.

@kshitij-v-mehta
Copy link
Collaborator Author

Closing this PR, as #272 replaces it

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants