GPU-Enhanced DFTB Metadynamics for Efficiently Predicting Free Energies of Biochemical Systems

Metadynamics calculations of large chemical systems with ab initio methods are computationallyprohibitive due to the extensive sampling required to simulate the large degrees of freedomin these systems. To address this computational bottleneck, we utilized a GPU-enhanced densityfunctional tight binding (DFTB) approach on a massively parallelized cloud computing platform toefficiently calculate the thermodynamics and metadynamics of biochemical systems. To first validateour approach, we calculated the free-energy surfaces of alanine dipeptide and showed that ourGPU-enhanced DFTB calculations qualitatively agree with computationally-intensive hybrid DFTbenchmarks, whereas classical force fields give significant errors. Most importantly, we show thatour GPU-accelerated DFTB calculations are significantly faster than previous approaches by up totwo orders of magnitude. To further extend our GPU-enhanced DFTB approach, we also carriedout a 10 ns metadynamics simulation of remdesivir, which is prohibitively out of reach for routineDFT-based metadynamics calculations. We find that the free-energy surfaces of remdesivir obtainedfrom DFTB and classical force fields differ significantly, where the latter overestimates the internalenergy contribution of high free-energy states. Taken together, our benchmark tests, analyses, andextensions to large biochemical systems highlight the use of GPU-enhanced DFTB simulations forefficiently predicting the free-energy surfaces/thermodynamics of large biochemical systems.

Automated summary

To overcome the computational challenges of large chemical systems, researchers used a GPU-enhanced DFTB approach on a massively parallelized cloud computing platform to efficiently calculate the thermodynamics and metadynamics of biochemical systems. The benchmark tests and extensions demonstrated that GPU-accelerated DFTB simulations are significantly faster and can predict the free-energy surfaces/thermodynamics of large biochemical systems effectively.