Bridging Quantum and Classical Computing in Drug Design: Architecture Principles for Improved Molecule Generation

TL;DR

This paper develops optimized hybrid quantum-classical neural network architectures for drug molecule generation, significantly improving performance and providing empirical guidelines for integrating quantum computing into pharmaceutical research.

Contribution

It introduces the BO-QGAN model with optimized architecture, offering the first empirical guidelines for hybrid quantum-classical models in drug discovery.

Findings

BO-QGAN achieves 2.27-fold higher DCS than previous benchmarks.

Layering multiple shallow quantum circuits enhances performance.

Classical architecture sensitivity is less above a minimum capacity.

Abstract

Hybrid quantum-classical machine learning offers a path to leverage noisy intermediate-scale quantum (NISQ) devices for drug discovery, but optimal model architectures remain unclear. We systematically optimize the quantum-classical bridge architecture of generative adversarial networks (GANs) for molecule discovery using multi-objective Bayesian optimization. Our optimized model (BO-QGAN) significantly improves performance, achieving a 2.27-fold higher Drug Candidate Score (DCS) than prior quantum-hybrid benchmarks and 2.21-fold higher than the classical baseline, while reducing parameter count by more than 60%. Key findings favor layering multiple (3-4) shallow (4-8 qubit) quantum circuits sequentially, while classical architecture shows less sensitivity above a minimum capacity. This work provides the first empirically-grounded architectural guidelines for hybrid models, enabling more effective integration of current quantum computers into pharmaceutical research pipelines.