An Artificial Intelligence-Informed Review of the Antioxidant and Anticancer Properties of Onion (Allium cepa) and Tomato (Solanum lycopersicum) in Prostate Cancer Prevention and Management: A Contemporary Review

Abstract

Prostate Cancer (PCa) is the second most common male cancer and fifth leading cause of cancer death, with 1.4 million new cases and 375,000 deaths annually. Therapies are limited by castration-resistant prostate cancer, multidrug resistance and dose-limiting toxicities. The oxidative stress-inflammatory axis driven by ROS-induced DNA damage, NF-κB activation and dysregulated androgen receptor signaling is central to PCa initiation, progression and metastasis. Dietary phytochemicals show chemopreventive promise, but translation is hindered by poor target characterization, limited bioavailability and lack of predictive prioritization frameworks. This review synthesizes current evidence on antioxidant and anticancer phytoconstituents from Allium cepa (onion) and Solanum lycopersicum (tomato), using AI-driven approaches molecular docking, network pharmacology, machine learning and deep learning for compound prioritization across three pillars: oxidative stress-inflammation-cancer modulation, phytochemical polypharmacology and predictive AI modeling. Quercetin activates the Nrf2/ARE pathway via Keap1 modification, while lycopene quenches singlet oxygen through its conjugated polyene chain. AI docking identified quercetin as a higher-affinity androgen-receptor binder than enzalutamide and machine learning models predicted antioxidant capacity from metabolomic profiles with strong cross-validation. However, clinical evidence remains equivocal, with meta-analyses showing no significant effect of lycopene on prostate-specific antigen levels. The review proposes a translational pathway addressing bioavailability optimization, biomarker refinement and natural product-specific AI model development.