Keywords
Digital intelligence
Nutrition and health product development
Data model
Submitted : 2025-12-16
Accepted : 2025-12-31
Published : 2026-01-15
Abstract
This study addresses common bottlenecks in the global biopharmaceutical industry, including slow R&D cycles, high compliance costs, and challenges in international market access. It proposes an evidence-based accelerated model for nutritional supplements that integrates digital intelligence, biotech innovation, and regulatory science. Centered on Australia's TGA pharmaceutical-grade regulatory framework, the model establishes an integrated global system encompassing R&D, efficacy validation, and international compliance. Leveraging a big data platform for active ingredients and an AI-driven formulation engine, it combines biological mechanism analysis with real-world evidence (RWE) validation. The architecture features standardized core modules and region-specific customization to accommodate global regulatory variations. A practical case study of WALVE Biotech's Revefore® ternary synergistic formulation demonstrates the model's effectiveness. Research confirms that this approach not only enhances profit margins per product unit but also enables rapid adaptation to global markets through dynamic integration with regional regulatory databases like FDA and EFSA. The model provides a high-standard, efficient R&D solution for the evidence-based nutrition industry, offering broad applicability across the sector.
References
Zhang J, Li Y, Wang H, 2023, End-to-end prediction of component-target-efficacy associations in nutraceuticals using a Transformer-based deep learning model. Nature Biotechnology, 41(8): 1123-1132.
Zhang H P, Zhang W Y, Li J, 2025, AI-driven full-chain R&D framework for probiotics: From strain screening to industrial transformation. Trends in Food Science & Technology, 101: 103892.
Jebb S A, Smith L, Brown C, 2024, Personalized vitamin D supplementation based on reinforcement learning integrating genetic and microbiome data. The Lancet Diabetes & Endocrinology, 12(5): 356-366.
Kan J T, Zhang Y, Liu X, 2020, XGBoost-based visual health score prediction for personalized lutein ester and zeaxanthin formulation. Frontiers in Nutrition, 7: 586432.
Kelleher R P, Walsh J M, Murphy E A, 2021, Real-world evidence for safety assessment of vitamin D supplements in older adults with comorbidities. New England Journal of Medicine, 385(12): 1103-1112.
Bowden L J, Chen M, Anderson K S, 2024, Real-world evidence to support TGA re-registration of nutraceuticals: A case study of NMN-based products. Drug Safety, 47(3): 289-300.
Tovey M G, Roberts C K, Patel S A, 2023, A quality grading system for real-world evidence in nutraceutical evaluation. European Journal of Nutrition, 62(4): 2185-2198.
Malin J L, Zhang L, Ahuja S, 2022, Federated learning and blockchain for cross-border sharing of real-world nutraceutical data. JAMA Network Open, 5(7): e2224561.
Kesselheim A S, Avorn J, Sarpatwari A, 2023, AI-driven regulatory decision support for nutraceuticals: A prospective evaluation. JAMA, 330(12): 1145-1153.
Schwartz L M, Woloshin S, Welch H G, 2022, A three-dimensional framework for integrating digital intelligence with regulatory science. Nature Reviews Drug Discovery, 21(9): 663-678.
Feihe Aiben R&D Team, Huawei Cloud, 2025, Regulatory-adaptive AI R&D model for functional nutraceuticals. China Food and Drug Administration, 31(2): 45-53.
Sharkey C M, Evans W N, 2021, Algorithmic fairness in nutraceutical regulatory science. Yale Journal of Health Policy, Law, and Ethics, 21(2): 317-348.