Predictive modeling of antioxidant synergies in fruits and vegetables: A multi-objective optimization approach for...

Complex interactions among phytochemicals in fruits and vegetables underpin their antioxidant benefits. Here, we introduce a multi-objective optimization framework to predict optimal functional-food combinations. Twelve key phytochemicals were evaluated at nine ratios using DPPH and ABTS assays, revealing ratio-dependent synergy or antagonism (e.g., β-carotene/epicatechin at 1:9 yielded -19.94 % antagonism versus +13.36 % synergy at 19:1 in ABTS). A Python-based prediction system minimizes calculated EC₅₀ values to infer optimal produce pairings. Validation against experimental DPPH synergy data showed a significant negative correlation (r = -0.69, p = 0.04), confirming the model's predictive accuracy. The ABTS model correlated more weakly (r = -0.55, p = 0.13), highlighting areas for refinement. This work offers a robust, data-driven approach for designing antioxidant-rich fruit and vegetable combinations, reducing experimental burden and informing evidence-based dietary and functional-food recommendations to enhance public health.