Quantum-Centric Alchemical Free Energy Calculations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bazayeva, Milana, Li, Zhen, Kaliakin, Danil, Liang, Fangchun, Shajan, Akhil, Das, Susanta, Merz Jr, Kenneth M.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908422382813184
author Bazayeva, Milana
Li, Zhen
Kaliakin, Danil
Liang, Fangchun
Shajan, Akhil
Das, Susanta
Merz Jr, Kenneth M.
author_facet Bazayeva, Milana
Li, Zhen
Kaliakin, Danil
Liang, Fangchun
Shajan, Akhil
Das, Susanta
Merz Jr, Kenneth M.
contents In the present work, we present a hybrid quantum-classical workflow aimed at improving the accuracy of alchemical free energy (AFE) predictions by incorporating configuration interaction (CI) simulations using the book-ending correction method. This approach applies the Multistate Bennett Acceptance Ratio (MBAR) over a coupling parameter λ to smoothly transition the system from molecular mechanics (MM) (λ = 0) to a quantum mechanics (QM) (λ = 1) description. The resulting correction is then applied to the classically (MM) computed AFE to account for the more accurate QM treatment. The standard book-ending procedure uses AMBER to simulate the MM region, and QUICK, AMBER's default QM engine, to handle the QM region with either the Hartree-Fock (HF) method or density functional theory (DFT). In this work, we introduce a novel interface to QUICK, via sander, that enables CI simulations, and can operate in two ways: A) via PySCF backend to perform full configuration interaction (FCI) using conventional computing resources, B) quantum-centric sample-based quantum diagonalization (SQD) workflow via Qiskit which leverages both quantum hardware and post-processing on conventional computing resources for CI simulations. In this workflow QUICK performs most steps of the calculations, but at user-defined intervals, it redirects the computation to either FCI or SQD backend to get the CI result. We computed the book-end corrections for the hydration free energy (HFE) of three small organic molecules (ammonia, methane, and water) to benchmark the proposed approach and demonstrate how quantum-computers can be used in AFE calculations. We believe that this approach can be scaled to more complex systems like drug-receptor interactions in future studies.
format Preprint
id arxiv_https___arxiv_org_abs_2506_20825
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Quantum-Centric Alchemical Free Energy Calculations
Bazayeva, Milana
Li, Zhen
Kaliakin, Danil
Liang, Fangchun
Shajan, Akhil
Das, Susanta
Merz Jr, Kenneth M.
Chemical Physics
Quantum Physics
In the present work, we present a hybrid quantum-classical workflow aimed at improving the accuracy of alchemical free energy (AFE) predictions by incorporating configuration interaction (CI) simulations using the book-ending correction method. This approach applies the Multistate Bennett Acceptance Ratio (MBAR) over a coupling parameter λ to smoothly transition the system from molecular mechanics (MM) (λ = 0) to a quantum mechanics (QM) (λ = 1) description. The resulting correction is then applied to the classically (MM) computed AFE to account for the more accurate QM treatment. The standard book-ending procedure uses AMBER to simulate the MM region, and QUICK, AMBER's default QM engine, to handle the QM region with either the Hartree-Fock (HF) method or density functional theory (DFT). In this work, we introduce a novel interface to QUICK, via sander, that enables CI simulations, and can operate in two ways: A) via PySCF backend to perform full configuration interaction (FCI) using conventional computing resources, B) quantum-centric sample-based quantum diagonalization (SQD) workflow via Qiskit which leverages both quantum hardware and post-processing on conventional computing resources for CI simulations. In this workflow QUICK performs most steps of the calculations, but at user-defined intervals, it redirects the computation to either FCI or SQD backend to get the CI result. We computed the book-end corrections for the hydration free energy (HFE) of three small organic molecules (ammonia, methane, and water) to benchmark the proposed approach and demonstrate how quantum-computers can be used in AFE calculations. We believe that this approach can be scaled to more complex systems like drug-receptor interactions in future studies.
title Quantum-Centric Alchemical Free Energy Calculations
topic Chemical Physics
Quantum Physics
url https://arxiv.org/abs/2506.20825