Exact and efficient calculation of derivatives of Lagrange multipliers for molecular dynamic simulations of biological molecules

TL;DR

This paper presents an exact and efficient method for calculating derivatives of Lagrange multipliers in constrained molecular dynamics, enabling the use of high-order integrators without computational bottlenecks.

Contribution

It introduces an analytical approach with linear scaling complexity for derivatives of constraint forces, facilitating high-order constrained molecular dynamics simulations.

Findings

Derivatives of constraint forces can be computed analytically with linear complexity.

The method ensures feasibility of high-order integrators in constrained MD.

Efficient calculation prevents bottlenecks in biological molecule simulations.

Abstract

In the simulation of biological molecules, it is customary to impose constraints on the fastest degrees of freedom to increase the time step. The evaluation of the involved constraint forces must be performed in an efficient manner, for otherwise it would be a bottleneck in the calculations; for this reason, linearly-scaling calculation methods have become widely used. If integrators of order higher than 2 (e.g. Gear predictor-corrector methods) are used to find the trajectories of atoms, the derivatives of the forces on atoms with respect to the time also need to be calculated, which includes the derivatives of constraint forces. In this letter we prove that such calculation can be analytically performed with linearly scaling numerical complexity (O(Nc), being Nc the number of constraints). This ensures the feasibility of constrained molecular dynamics calculations with high-order integrators.