PerspectiveFeatureHigh-throughput molecular dynamics: the powerful new tool for drug discovery
Highlights
► Molecular dynamics (MD) is able to provide quantitative insight into the kinetics of drug-ligand binding processes. ► MD cost substantially reduced by new developments in computing hardware. ► Markov state modelling using moves MD simulation from a high performance to high-throughput activity. ► High-throughput MD now a valuable tool in computational drug design.
Section snippets
Technology of MD simulation
MD modelling of biomolecules typically treats each atom of the solvent and solute as separate point particle. A force-field, parametrised to capture the chemical properties of the environment of each type of particle governs the evolution of the system, which proceeds according to Newtonian dynamics in a stepwise manner. At each step of the simulation, the net force on each particle is calculated and the particles’ positions updated.
To correctly capture the dynamics of the system the timestep
Exploiting high-throughput MD
The desire to access long timescales within a reasonable time has led to the development of a variety of enhanced sampling techniques, which typically impose constraints or biases to accelerate the evolution of a system. These methods require some a priori knowledge about the system so that a reaction coordinate or order parameter can be defined to guide the enhanced sampling. The evolution of these techniques has been motivated by the need to extract the maximum sampling from single long
Concluding remarks
In this perspective we argue that the reduction in cost of GPUs, together with high-throughput MD protocols can provide a way to resolve the physical chemistry of molecular recognition, for instance using reasonably sized GPU clusters and high-throughput protocols. MD simulation should move from high performance computing, with its focus on maximising length of a small number of individual simulations, to high-throughput, where large ensembles of independent, ‘long enough’ simulations may be
Conflict of interest
The authors declare a financial interest in Acellera Ltd.
Acknowledgement
The authors thank David Soriano for a critical reading of the manuscript.
References (31)
- et al.
Reaching biological timescales with all-atom molecular dynamics simulations
Curr. Opin. Pharmacol.
(2010) The impact of accelerator processors for high-throughput molecular modeling and simulation
Drug Discov. Today
(2008)How robust are protein folding simulations with respect to force field parameterization?
Biophys. J.
(2011)Using generalized ensemble simulations and Markov state models to identify conformational states
Methods
(2009)Mechanisms of protein-ligand association and its modulation by protein mutations
Biophys. J.
(2011)ACEMD: accelerated molecular dynamics simulations in the microseconds timescale
J. Chem. Theory Comput.
(2009)Binding estimation after refinement, a new automated procedure for the refinement and rescoring of docked ligands in virtual screening
Chem. Biol. Drug. Dis.
(2009)Virtual screening: an endless staircase?
Nat. Rev. Drug. Dis.
(2010)A role for both conformational selection and induced fit in ligand binding by the LAO protein
PLoS Comp. Biol.
(2010)Cramming more components on to integrated circuits
Electronics
(1965)