Graph neural network modeling of receptor interaction kinetics from single-molecule imaging data
The paper develops Deep-FISIK, a graph neural network, to predict receptor interaction kinetics from single-molecule imaging data without explicit tracking.