Stefan C. Pate

This paper develops reaction rules informed by enzyme mechanisms to predict viable novel enzymatic reactions, achieving high coverage and precise atom identification.

RC-GNN predicts enzyme-reaction pairs using amino acid sequences and reaction stoichiometry, achieving high ROC-AUC scores on diverse datasets.

This paper shows how neuromodulators alter neural network behavior by shifting activity flows through hyperchannels, enhancing computational flexibility in a recurrent neural network model.

The paper identifies two distinct GPe-PV neuron pathways in mice and shows they contribute to different Parkinsonian symptoms through differential electrophysiological responses to dopamine depletion.