Convolution graph model for embedding molecular structure

[ph-mehdi]

hi I want to design a torsional diagram model that returns an embedded vector from a molecular structure. This convolutional graph model has graph-conv, graph-pooling and graph-gather layers. I may also give the taken vector to another model such as lstm. Is this feasible? Can you guide me?

[mufeili]

I believe it's feasible to embedding a molecule into a vector with the convolutional graph model. You can also use recurrent neural networks like GRUs or LSTMs in computing the molecule embeddings. For an example, see this code snippet.

[github-actions[bot]]

This issue has been automatically marked as stale due to lack of activity. It will be closed if no further activity occurs. Thank you

[github-actions[bot]]

This issue is closed due to lack of activity. Feel free to reopen it if you still have questions.