Hybrid experimental-QSAR-artificial intelligence framework for chronic ecological risk assessment of emerging pollutants in coastal environments - PubMed
8 hours ago
- #coastal pollutants
- #ecological risk assessment
- #QSAR-AI framework
- A hybrid framework combines experimental data, QSAR tools, and AI to assess chronic ecological risks of emerging pollutants in coastal environments.
- Experimental results identified acetaminophen and caffeine as high-risk substances, with diclofenac posing negligible risk.
- QSAR screening highlighted caffeine, losartan, and benzoylecgonine as priority pollutants.
- VEGA analysis showed reliable predictions for about 65-70% of compound-taxon combinations.
- AI model TRIDENT moderated extreme QSAR outputs, confirming high risks for caffeine, acetaminophen, and losartan.
- Algae were the most sensitive taxonomic group, followed by crustaceans and fish.
- The framework enhances ecological risk assessment in data-limited tropical coastal areas and is internationally transferable.