Jul 2026· IEEE International Conference on Circuits and Systems for Communications· pp. 1-7· 0 citations· 25 references
Abstract
Generative models for de novo molecular design have become a cornerstone of computational drug discovery, yet existing hybrid quantum-classical approaches remain confined to small-molecule benchmarks such as QM9 ($\boldsymbol{\leq} \mathbf{9}$ heavy atoms). This work introduces BO-QGAN+, a Bayesian-optimised hybrid quantum generative adversarial network that, for the first time, demonstrates competitive molecular generation on the ZINC-250k benchmark, accommodating molecules of up to 38 heavy atoms. Our primary methodological novelty lies in a multicircuit quantum ensemble generator featuring data reuploading variational circuits, which enables scaling to the substantially larger molecular graphs present in ZINC-250k. This generator is coupled with a Wasserstein critic stabilised by gradient penalty (WGAN-GP) and a graph-convolutional discriminator with spectral normalisation. On a 5,000-molecule ZINC-250k subset, BO-QGAN+ achieves the best Drug Candidate Score (DCS) of 1.244 with 94% molecular validity, a QED of 0.41, normalised logP of 0.71, and normalised SA of 0.43 within 700 training epochs These results are obtained with only 16 quantum parameters alongside 2.03 M classical parameters, demonstrating a parameter efficiency ratio exceeding 127,000:1. At epoch 700, the model further achieves Uniqueness of 0.97, Novelty of 0.84, and Internal Diversity of 0.79, confirming that high validity does not come at the expense of generation diversity. Our findings establish that hybrid quantum generators can operate effectively at molecular scales well beyond prior quantum GAN benchmarks, Key limitations include the use of classical quantum simulation rather than physical NISQ hardware, and evaluation on a 5,000-molecule subset. Nevertheless, our findings open a practical pathway toward quantum-assisted drug discovery pipelines on near-term quantum hardware.
This work constitutes a reproducible baseline for noise-aware quantum machine learning on pharmaceutical data and shows that hardware-efficient entanglement structures—specifically alternating cost and mixer layers—provide a systematic advantage over purely rotational or diagonal encodings under realistic noise conditions.
Rania Derouich, N. Mathlouthi· bioRxiv· 0 citations
Overall, the results show that quantum and hybrid quantum-classical generative models can learn non-trivial discrete probability distributions but that their effectiveness depends strongly on the selected quantum model, ansatz, and training objective.
This study integrated the conditional variational autoencoder with the Wasserstein generative adversarial network and effectively applied this hybrid architecture to molecular generation tasks, introducing a molecular generation framework with conditional generation capabilities known as CCVAN.
Molecular generation is a core task in drug discovery. Although existing deep generative models can produce valid molecular structures, they lack precise control over molecular properties such as lipophilicity (logP) and drug-likeness (QED). This paper proposes Property-Regularized Graph Variational Autoencoder (PR-GVAE), a conditional graph VAE that introduces Feature-wise Linear Modulation (FiLM) conditioning in the encoder and a dual-branch fusion architecture in the decoder for precise property control. Experiments on a filtered subset of ZINC-20 (approximately 52,000 molecules, at most 20 heavy atoms) across five models spanning graph-based, sequence-based, and fragment-assembly paradigms demonstrate that PRGVAE achieves a Condition Satisfaction Rate (CSR) for both properties of 78.3%, outperforming the unconditional Vanilla GraphVAE by 2.0× (39.4%) and the sequence-based SMILES C-VAE by 40 percentage points (38.3%). Ablation experiments reveal that decoder conditioning is the core driver of property control (removing it reduces CSR-Both by 26.6%), while encoder FiLM conditioning and dual-branch fusion exhibit additive contributions (simultaneous removal reduces CSR-Both by 30.8%). Visualization confirms that PR-GVAE generates structurally diverse drug-like molecules, providing an effective solution for on-demand molecular design.
Lu Yu, Tianyu Han, Shuyue Men et al.· 2026 IEEE 27th China Confere...· 0 citations
Deep generative models have become a leading approach for designing therapeutic molecules, yet efficiently exploring vast biomolecular sequence spaces remains difficult, particularly for targets with limited training data. The prior distribution that seeds a generative model shapes which regions of sequence space it explores, and recent work suggests that non-classical distributions sampled from quantum processors can serve as a structured alternative to the factorised Gaussian priors used by default. Whether such priors help on complex biological design tasks has been largely untested. Here we present what is, to our knowledge, the first end-to-end hybrid quantum-classical pipeline for de novo design of MHC class I-binding peptides, coupling a generative adversarial network (GAN) to latent vectors sampled from a real photonic quantum processor. Tested in silico across 131 HLA alleles, quantum-derived priors increased the yield of predicted strong binders, with the largest relative gains for understudied alleles where classical baselines perform worst. We selected three understudied alleles for further evaluation, finding that large gains coincided with broader sequence exploration at non-anchor positions while anchor specificity was preserved. On these three alleles, we validated the designs in vitro using peptide-MHC stability ELISAs, confirming that quantum-designed peptides are potent stabilisers of peptide-MHC class I complexes. These results establish structured, hardware-realisable non-classical priors as a useful inductive bias for generative peptide design, with direct relevance to personalised immunotherapies and vaccines.
E. S. Engdal, Jonathan Funk, O. Bacarreza et al.· bioRxiv· 0 citations
