Answers are grounded in this paper and cite their sources.
Plant genomics with a hybrid classical-quantum model Artificial intelligence and quantum computing for the digital design of plant genomes Italian patent application no. 102026000034324, filed on 18 November 2026. Inventor, applicant and sole rights holder: Paolo Ciccolella. This invention proposes combining artificial intelligence and quantum computing to explore possible combinations of plant DNA and identify those worth investigating further. For businesses interested in plant genomics, the idea is to provide early guidance for computational design: turning a vast number of possibilities into selected, evaluated and documented digital candidates. The challenge is easy to picture. If DNA is a large instruction book, studying alternative versions of some passages means comparing many combinations. Three alternatives for each of twenty regions already produce more than three billion possibilities. The method tackles this complexity by dividing the work between a classical computer, two AI models and a quantum search component. HOW THE HYBRID MODEL WORKS The idea is to choose one alternative for each DNA region being studied, put the choices together and evaluate the resulting combination. Everything happens on a computer. The process follows six steps. • Prepare the data. The computer starts with a digital version of the plant’s DNA, called a reference genome. It identifies the regions to study and organises the available alternative sequences for each one, recording where they come from. • Give each alternative an initial score. The first AI model examines each alternative separately. Having learned from other cases, it can provide guidance before any complete combination in the new case has been evaluated. The computer also measures similarity between the ends of sequences that will be placed next to one another. These scores and measurements define the selection problem. [B12] • Search for combinations. The quantum component receives this problem expressed in numbers. It uses QAOA, an algorithm whose role is to propose combinations while seeking to favour those with better scores. Qubits represent the possible choices; the DNA sequences remain on the classical computer, which sets and updates the algorithm’s parameters. [B11] • Evaluate the whole combination. The computer checks that each combination selects exactly one alternative per region and assembles valid choices into a digital candidate. The second AI model evaluates the complete candidate, including relationships between its parts that the first AI model could not observe when examining them separately. • Guide the next search. The system compares the overall evaluation with the score predicted at the start and uses the difference to update the search. When only a few results are available, it limits the correction’s influence to avoid relying too heavily on sparse information. • Return verifiable proposals. For each selected candidate, the system retains the chosen sequences, their provenance and the evaluations performed. It also records a digital fingerprint, a code used to check the content. Researchers can reconstruct the proposal and understand how it was selected. If necessary data are missing, the system flags this. A concrete example: the patent uses the digital reference for Micro-Tom tomato and considers three regions, each with three alternatives created for the test. There are 3 × 3 × 3 = 27 possible combinations. The search uses a nine-qubit quantum circuit simulated on a classical computer. This tests the computational method without producing plants. [B14] INDUSTRIAL APPLICATIONS The patent describes the production of verifiable digital representations of candidate genomes. Its industrial role lies in designing and selecting hypotheses, ahead of experimental work. Applications across sectors require data and models suited to the objectives under study. • NGT and assisted evolution techniques. NGTs, or new genomic techniques, include approaches known in Italy as TEA, meaning assisted evolution techniques, which enable targeted changes to genetic material. The proposed method could support research by comparing genomic hypotheses before experimental work. For seed and biotechnology companies, the prospect is to build a reasoned set of digital candidates for further investigation within NGT/TEA programmes. • Seed industry and variety development. For seed companies and research teams, the method could support the preliminary selection of combinations for further study. Research objectives could include drought or salinity tolerance, disease resistance and yield stability. Its contribution would be to establish research priorities while preserving the reasoning behind them. • Fresh produce and food processing. Computational design could focus on commercial characteristics such as product quality, shelf life and processing suitability. Tomato, pepper and aubergine are examples of supply chains that future projects could address, connecting growers’ and processors’ needs with the criteria used to evaluate candidates. • Nutraceuticals and plant ingredients. One possible development concerns plants of interest for functional ingredients, extracts and bioactive compounds. The method could explore genomic combinations in programmes focused, for example, on pigments or plant metabolites, providing documented candidates to guide subsequent analysis. • Plant biotechnology and industrial raw materials. For crops intended for oils, fibres or biomass, the method could help compare genomic hypotheses against supply chain objectives. Reconstructible digital records would support information exchange between teams developing the models and those planning subsequent experimental work. • Bioinformatics and computing centres. The most direct application is within the information systems of laboratories, computing centres and plant genomics businesses. The process could underpin software modules or analytical services that explore combinations and return candidates with scores and execution logs, integrating quantum search into the workflow. REFERENCES References from the patent application. Numbering B1–B14 matches the filed description. [B1] EP 4 544 555 A1, Illumina, Inc. “Generating and implementing a structural variation graph genome”. Priorità / Priority: 2022-06-27. Pubblicazione / Publication: 2025-04-30. [B2] J. Cudby, J. Bonfield, C. Zhou, R. Durbin, S. Strelchuk. “Pangenome-guided sequence assembly via binary optimisation”. 2025-08-11. arXiv:2508.08200. https://arxiv.org/abs/2508.08200 [B3] J. Cudby, S. Strelchuk. “Nonvariational quantum optimisation approaches to pangenome-guided sequence assembly”. 2026-04-07. arXiv:2604.06106. https://arxiv.org/abs/2604.06106 [B4] K. Kitai et al. “Expanding the horizon of automated metamaterials discovery via quantum annealing”. 2019-02-18. arXiv:1902.06573. https://arxiv.org/abs/1902.06573 [B5] R. Tamura et al. “Black-box optimization using factorization and Ising machines”. 2025-07-24. arXiv:2507.18003. https://arxiv.org/abs/2507.18003 [B6] WO 2024/211046 A2, Cornell University. “Molecular design with quantum-computing-based machine learning and optimization”. Priorità / Priority: 2023-03-07. Pubblicazione / Publication: 2024-10-10. [B7] S. Kikuchi, S. Tanaka. “High-Order Epistasis Detection Using Factorization Machine with Quadratic Optimization Annealing and MDR-Based Evaluation”. 2026-01-05. arXiv:2601.01860. https://arxiv.org/abs/2601.01860 [B8] E. Alessandroni, S. Ramos-Calderer, I. Roth, E. Traversi, L. Aolita. “Alleviating the quantum Big-M problem”. 2023-07-19. arXiv:2307.10379. https://arxiv.org/abs/2307.10379 [B9] M. Ayodele. “Penalty Weights in QUBO Formulations: Permutation Problems”. 2022-06-20. arXiv:2206.11040. https://arxiv.org/abs/2206.11040 [B10] E. Doucet et al. “Thermodynamic significance of QUBO encoding on quantum annealers”. 2026-01-07. arXiv:2601.04402. https://arxiv.org/abs/2601.04402 [B11] E. Farhi, J. Goldstone, S. Gutmann. “A Quantum Approximate Optimization Algorithm”. 2014-11-14. arXiv:1411.4028. https://arxiv.org/abs/1411.4028 [B12] A. Lucas. “Ising formulations of many NP problems”. Frontiers in Physics, 2014, vol. 2, art. 5. https://doi.org/10.3389/fphy.2014.00005 [B13] J. Preskill. “Quantum Computing in the NISQ era and beyond”. Quantum, 2018, vol. 2, p. 79. https://doi.org/10.22331/q-2018-08-06-79 [B14] National Center for Biotechnology Information. Assemblaggio / Assembly: GCF_036512215.1. Rilascio di annotazione / Annotation release: GCF_036512215.1-RS_2024_09. https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_036512215.1/ Patent Application Title: HYBRID CLASSICAL-QUANTUM PROCESS FOR THE IN SILICO RECOMBINATION OF PLANT GENOMES WITH EVALUATION BY MEANS OF ARTIFICIAL INTELLIGENCE • Inventor and owner: Paolo Ciccolella — inventor, applicant and sole rights holder • Application: 102026000034324 • Filing date: 18 November 2026 (2026-11-18) • Office: Ufficio Italiano Brevetti e Marchi (UIBM) — Italy • Status: Industrial invention patent application filed
No takes yet. Share an insight, caveat, or question.
Paolo Ciccolella (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: