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].