Drug Design in the Exascale Era: A Perspective from Massively Parallel QM/MM Simulations

Abstract

The initial phases of drug discovery – in silico drug design – could benefit from first principle Quantum Mechanics/Molecular Mechanics (QM/MM) molecular dynamics (MD) simulations in explicit solvent, yet many applications are currently limited by the short time scales that this approach can cover. Developing scalable first principle QM/MM MD interfaces fully exploiting current exascale machines – so far an unmet and crucial goal – will help overcome this problem, opening the way to the study of the thermodynamics and kinetics of ligand binding to protein with first principle accuracy. Here, taking two relevant case studies involving the interactions of ligands with rather large enzymes, we showcase the use of our recently developed massively scalable Multiscale Modeling in Computational Chemistry (MiMiC) QM/MM framework (currently using DFT to describe the QM region) to investigate reactions and ligand binding in enzymes of pharmacological relevance. We also demonstrate for the first time strong scaling of MiMiC-QM/MM MD simulations with parallel efficiency of ∼70% up to >80,000 cores. Thus, among many others, the MiMiC interface represents a promising candidate toward exascale applications by combining machine learning with statistical mechanics based algorithms tailored for exascale supercomputers.

Perspectives on Ligand/Protein Binding Kinetics Simulations: Force Fields, Machine Learning, Sampling, and User-Friendliness

“…The hardware and software improvements—also considering the emerging era of quantum computing—together with the continuously evolving ML techniques will undoubtedly play a leading role in the coming years, hopefully making this kind of calculations less demanding and useful for ligand database screening protocols. In this regard, one could expect that even quantum mechanical calculations—now considered unfeasible for ligand binding studies 203 —might at least integrate the atomistic level description of LPB. In the present article, we have discussed three macro areas that could be further improved, which are molecular properties parametrization, sampling, and data openness.…”

All-Atom Biomolecular Simulation in the Exascale Era

“…Further, many phenomena in biology, including enzyme catalysis, proton pumping, and electron transfer reactions, require a quantum mechanical treatment necessitating the use of QM/MM methods or quantum-based surrogate models . Some of them have been massively parallelized in order to exploit exascale resources for modeling large systems. , Finally, while extensive work has gone into optimizing molecular force fields that mimic the basic interactions between atoms, it is now accepted that electronic polarization and many-body dispersion forces can contribute substantially and can even qualitatively alter the equilibrium structures and self-assembly dynamics . These effects will move to the fore and require additional capabilities as we attempt to model large assemblies at an accurate level …”

A Vision for the Future of Multiscale Modeling

“…Additionally, all of these processes occur over tens to hundreds of Å, a scale that significantly challenges the actual multiscale implementations. Furthermore, some of these phenomena still extend beyond current computational power as they would require QM simulations to span in the ns time scale, ,,, while the MM part should account for heavy atom motion occurring in the μs or ms time scales. In this regard, some promising results have been recently achieved using MLIPs techniques, as discussed in Sec…”