6.2 Integrators and autotuning
Over the past two decades, MSM has emerged into a promising tool to build in-silico predictive models by systematically integrating knowledge from the tissue, cellular, and molecular level. Depending on the scale of interest, governing equations in each scale of the MSM approaches may fall into two categories, ordinary differential equation-based, and partial differential equation-based approaches. Examples include molecular dynamics [21], coarse-grained mesoscale models [71], lattice Boltzmann methods [43], immersed boundary methods [138], as well as classical finite element approaches [139]. ML-based methods can speed up, optimize, and autotune several of the existing solvers for multiphysics simulations [140, 141].