Systems Biology of Cell Behavior: Robotics, Machine Learning, Microfluidics, and Multiscale (INVITED SPEAKER)

Cells respond to numerous stimuli in space and time. We developed pairwise agonist scanning (PAS) using 384-well plate robotic assay to create high-dimensional and dynamic response data for neural network (NN) training. Lattice Kinetic Monte Carlo (LKMC) then allows for patient-specific simulation of large cell ensembles (>10,000 cells) in the presence of diffusion, convection, and chemical gradients. We apply this approach to predict human blood clotting and bleeding in response to diverse drug treatments using microfluidic assay conditions. This approach then becomes patient-specific and multiscale with 3D medical imaging of the coronary vasculature and individual NN models of platelet function.