Artificial intelligence-based modeling of oxidative stress biomarkers: a comparison with the integrated biomarker response (IBR) approach
EXPERT SYSTEMS WITH APPLICATIONS, cilt.333, 2027 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 333
- Basım Tarihi: 2027
- Doi Numarası: 10.1016/j.eswa.2026.134143
- Dergi Adı: EXPERT SYSTEMS WITH APPLICATIONS
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Applied Science & Technology Source, Compendex, INSPEC, Public Affairs Index, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO), Technology Collection (ProQuest)
- Gazi Üniversitesi Adresli: Evet
Özet
Oxidative stress biomarkers are widely used in ecotoxicological studies to evaluate the biological effects of environmental contaminants in aquatic organisms. The Integrated Biomarker Response (IBR) method is commonly applied to summarize multiple biomarker responses into a single index reflecting organism stress levels. However, the interpretation of complex biomarker datasets remains challenging, particularly when multiple biochemical indicators interact simultaneously. In this study, artificial intelligence-based modeling approaches were applied to analyze oxidative stress responses in freshwater mussels (Unio delicatus) exposed to pesticide contamination. The dataset consisted of biomarker measurements including total antioxidant status (TAS), total oxidant status (TOS), malondialdehyde (MDA), reduced glutathione (GSH), superoxide dismutase (SOD), and catalase (CAT). Multiple machine learning algorithms were evaluated to model the integrated biomarker response index. Model performance was assessed using the coefficient of determination (R2), mean absolute error (MAE), and root mean squared error (RMSE). Among the tested algorithms, regularized linear models demonstrated the highest predictive accuracy. In particular, ElasticNet and Lasso models achieved the best performance with R2 values of approximately 0.987. Feature importance and SHAP-based interpretability analyses further revealed the relative contributions of individual biomarkers to the integrated stress response. These suggest indicate that machine learning approaches may help explore relationships among oxidative stress biomarkers within the analyzed dataset and may provide complementary analytical tools for interpreting ecotoxicological data. The integration of artificial intelligence methods with traditional biomarker assessment techniques may offer new opportunities for improving environmental risk assessment.