Efficient Hamiltonian Simulation: A Utility Scale Perspective for Covalent Inhibitor Reactivity Prediction

TL;DR

This paper presents methods to significantly reduce quantum circuit depth for simulating complex molecular Hamiltonians, enabling more feasible quantum chemistry calculations on current noisy quantum hardware for drug development.

Contribution

It introduces a combination of Hamiltonian truncation, Clifford Decomposition, and optimized transpilation techniques to achieve substantial circuit depth reductions in quantum simulations of pharmaceutically relevant molecules.

Findings

Up to 28.5-fold circuit depth reduction with all-to-all connectivity.

Up to 15.5-fold reduction on IBMQ's Heron architecture.

Successful execution of large-scale Hamiltonian dynamics simulations on current hardware.

Abstract

Quantum computing applications in the noisy intermediate-scale quantum (NISQ) era require algorithms that can generate shallower circuits feasible for today's quantum systems. This is particularly challenging for quantum chemistry applications due to the inherent complexity of molecular systems. Working with pharmaceutically relevant molecules containing sulfonyl fluoride ($SO_2F$) warheads used in targeted covalent drug development, we combine Hamiltonian terms truncation, Clifford Decomposition and Transformation (CDAT), and optimized transpilation techniques to achieve up to a 28.5-fold reduction in circuit depth when assuming all-to-all connectivity of quantum hardware. When employed on IBMQ's Heron architecture, we demonstrate up to a 15.5-fold reduction. Through these methods, we reduced circuit depths to 1330 gates for 8-qubit Hamiltonian dynamics simulations. Using middleware solutions for circuit decomposition, we successfully executed sub-circuits with depths up to 371 gates containing 216 2-qubit gates, representing one of the largest electronic structure Hamiltonian dynamics calculations implemented on current quantum hardware. The systematic circuit reduction approach shows promise for scaling to larger active spaces, while maintaining sufficient accuracy for molecular reactivity predictions using the Quantum-Centric Data-Driven R&D framework. This work highlights practical methods for exploring commercially relevant chemistry problems on quantum hardware through Hamiltonian simulation, with direct applications to pharmaceutical drug development.