Abstract:Cadmium is a prominent toxic heavy metal that contaminates both terrestrial and aquatic environments. Owing to its high biological half-life and low excretion rates, cadmium causes a variety of adverse biological outcomes. Adverse outcome pathway (AOP) networks were envisioned to systematically capture toxicological information to enable risk assessment and chemical regulation. Here, we leveraged AOP-Wiki and integrated heterogeneous data from four other exposome-relevant resources to build the first AOP network relevant for inorganic cadmium-induced toxicity. From AOP-Wiki, we filtered 309 high confidence AOPs, identified 312 key events (KEs) associated with inorganic cadmium from five exposome-relevant databases using a data-centric approach, and thereafter, curated 30 cadmium relevant AOPs (cadmium-AOPs). By constructing the undirected AOP network, we identified a large connected component of 18 cadmium-AOPs. Further, we analyzed the directed network of 59 KEs and 82 key event relationships (KERs) in the largest component using graph-theoretic approaches. Subsequently, we mined published literature using artificial intelligence-based tools to provide auxiliary evidence of cadmium association for all KEs in the largest component. Finally, we performed case studies to verify the rationality of cadmium-induced toxicity in humans and aquatic species. Overall, cadmium-AOP network constructed in this study will aid ongoing research in systems toxicology and chemical exposome.

An integrative data-centric approach to derivation and characterization of an adverse outcome pathway network for cadmium-induced toxicity