Detecting the bioaccumulation patterns of chemicals through data-driven approaches

Area of Science:

• Environmental Chemistry
• Ecotoxicology
• Aquatic Toxicology

Background:

• Bioaccumulation of organic chemicals in aquatic ecosystems is a significant environmental concern.
• Understanding the relationship between different bioaccumulation metrics is vital for risk assessment.
• Existing metrics like octanol-water partitioning coefficient (KOW) and bioconcentration factor (BCF) are often used to predict biomagnification factor (BMF), but their efficacy is debated.

Purpose of the Study:

• To investigate bioaccumulation patterns of 168 organic chemicals in fish.
• To analyze the relationships between BCF, BMF, and KOW.
• To evaluate the effectiveness of KOW and BCF in predicting dietary bioaccumulation and identifying structure-related patterns.

Main Methods:

• Comparative analysis of bioconcentration factor (BCF), biomagnification factor (BMF), and octanol-water partitioning coefficient (KOW) for 168 organic chemicals.
• Linear regression analysis to establish relationships between logBMF, logBCF, and logKOW.
• Application of self-organizing maps and canonical correlation analysis to explore structure-related patterns in deviations from linear relationships.

Main Results:

• Linear relationships were observed between logBMF and logKOW (up to logKOW ≈ 4) and between logBMF and logBCF (up to logBCF ≈ 5).
• Approximately 10% of compounds deviated from these linear BCF-BMF relationships, with structure-related patterns identified.
• KOW- and BCF-based thresholds showed high false positive rates (47%-91% for KOW) and moderate false negative rates (up to 5% for BCF) for detecting dietary bioaccumulation.

Conclusions:

• KOW and BCF are insufficient predictors for compounds that bioaccumulate through diet.
• Experimental BMF is essential for accurate hazard assessment practices.
• Development of computational tools for predicting BMF is recommended.