Molecular dynamics (MD)

Molecular dynamics is a computational method that simulates how atoms and molecules move over time, by numerically solving the classical equations of motion for a system given a model of the forces acting between its particles.

A "force field" defines the interaction energies between atoms, and the simulation advances in small discrete time steps, computing forces and updating each particle's position and velocity at every step to produce a trajectory that is then analyzed for properties like diffusion, binding, or folding. This trades quantum-level accuracy for speed compared with density functional theory, which calculates electronic structure directly from quantum mechanics at far greater computational cost, MD's empirical force fields let it simulate systems of thousands to millions of atoms over much longer timescales.

Molecular dynamics is used throughout drug discovery, materials science, and protein-folding research, and it depends heavily on high-performance computing and GPU acceleration to be practical at useful scale, often run as one stage in a larger scientific workflow. Pitfalls include results that are only as good as the chosen force field's accuracy for the system being modeled, a time step too large to capture fast physical processes, and rare but important events, like a slow conformational change, that may require specialized sampling techniques to observe at all within a feasible simulation length.