MD Simulations
EXPLORATORYMolecular Dynamics (MD) simulations model how proteins move, fold, and interact with drug molecules over time. Each simulation runs on GPU hardware using OpenMM, tracking every atom at femtosecond resolution.
▶How does MD simulation setup work?
Each row represents a protein (or protein-drug complex) simulated under physiological conditions: 310 K (body temperature), explicit water solvent (TIP3P), 150 mM NaCl, periodic boundary conditions. Simulation types include SMN oligomerization, hnRNP A1–ISS-N1 binding, risdiplam mechanism, NCALD calcium dynamics, PLS3 actin bundling, and SMN–Gemin2 stability.
Key metrics. RMSD— < 0.6 A = stable (green), 0.6–1.0 A = moderate (amber), > 1.0 A = unstable (red). Binding energy— < −7 kcal/mol = strong drug binding. Contact persistence — percent of time the drug maintains key interactions.
DiffDock predicts a static binding pose; MD tests whether that pose is dynamically stable. A drug that stays bound for 100 ns is a much stronger candidate than one that dissociates after 10 ns. GPU hours estimate compute cost on a single NVIDIA A100.