Toxic effects of environmental biotoxin okadaic acid by network toxicology analysis and deep learning prediction

The study aims to promote a network toxicology and deep learning strategy to efficiently investigate the underlying neurotoxicity molecular mechanisms of okadaic acid (OA) , which is a typical representative of diarrhetic shellfish poisoning in bio-environmental or food chain system. 95 hub targets associated with OA-related diarrhea, and neurotoxicity were identified using K means algorithm of network toxicology strategy at the macro level. More specifically, the key target AKT1 was identified using DeepPurpose algorithm of deep learning strategy at the micro level. The synergistic integration of network toxicology and deep learning approaches enables multidimensional complementarity across systems biology and molecular interaction levels; the former constructs global toxicity networks, while the latter elucidates key target mechanisms. This multi-scale approach enhances study efficiency and mechanistic precision. Further on, molecular docking and bio-layer interferometry were conducted to confirm the binding between AKT1 and OA (INTERACTION_ENERGY =56.99 kcal/mol, K_D=6.61E-11 M). This research provides a theoretical and correlational basis of OA-induced diarrhea-related brain injury, as well as establishing a decision-making for the treatment of bio-environmental or foodborne OA exposure.