Theory and applications of the generalized born solvation model in macromolecular simulations

Key Points

• To evaluate the implementation, accuracy, and computational efficiency of Generalized Born implicit solvent models compared to explicit solvent simulations for biomacromolecules.
• Implemented and parallelized the Generalized Born (GB) model and a surface-area-based hydrophobic term within the AMBER molecular modeling suite.
• Performed comparative simulations between GB implicit solvent and explicit solvent systems using a 10 base-pair DNA oligomer and the 108-residue protein thioredoxin.
• Adjusted model parameters governing internal NH···O and NH···N hydrogen bond strengths and evaluated energy stability alongside computational timing.
• Parameter adjustments that slightly strengthened internal NH···O and NH···N hydrogen bonds brought GB simulation structures into close agreement with explicit solvent trajectories.
• GB models maintained conformational stability for both the DNA oligomer and the thioredoxin protein without explicitly simulating surrounding water molecules.
• Computational timing and energy stability benchmarks confirmed the scalability of the implementation for extended simulations of larger macromolecular systems.