Using pretrained graph neural networks with token mixers as geometric featurizers for conformational dynamics

Open in new window