Analytic energy, gradient, and hessian of electrostatic embedding QM/MM based on electrostatic potential fitted atomic charges scaling linearly with the MM subsystem size
Miquel Huix-Rotllant, Nicolas Ferr\'e

TL;DR
This paper introduces an improved electrostatic embedding QM/MM method using a new atomic charge operator that conserves total charge and includes grid derivatives, enabling efficient and accurate energy, gradient, and Hessian calculations that scale linearly with system size.
Contribution
The authors develop a new ESPF atomic charge operator that addresses previous limitations, ensuring charge conservation and including grid derivatives with minimal computational cost.
Findings
The new charge operator conserves total charge and improves translational invariance.
Analytic expressions for energy, gradient, and Hessian scale linearly with system size.
Application to cryptochrome demonstrates practical effectiveness.
Abstract
Electrostatic potential fitting method (ESPF) is a powerful way of defining atomic charges derived from quantum density matrices fitted to reproduce a quantum mechanical charge distribution in the presence of an external electrostatic potential. These can be used in the Hamiltonian to define a robust and efficient electrostatic embedding QM/MM method. The original formulation of ESPF QM/MM contained two main approximations, namely, the neglect of grid derivatives and the non-conservation of the total QM charge. Here, we present a new ESPF atomic charge operator which solves these drawbacks at virtually no extra computational cost. The new charge operators employ atom-centered grids and conserve the total charge when traced with the density matrix. We present an efficient and easy-to-implement analytic form for the energy, gradient, and hessian that scale linearly with the MM subsystem…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
