Predicting Antibiotic Resistance and Assessing the Risk Burden from Antibiotics: A Holistic Modeling Framework in a...

Area of Science:

• Environmental Science
• Microbiology
• Computational Modeling

Background:

• Antimicrobial resistance (AMR) in aquatic ecosystems poses significant risks.
• Effective prediction of AMR hotspots is crucial for risk management.
• Tropical reservoirs are important aquatic environments requiring AMR monitoring.

Purpose of the Study:

• To develop and validate an integrated modeling framework for predicting the spatiotemporal abundance of antibiotics, indicator bacteria, and antibiotic-resistant bacteria (ARB).
• To assess the potential AMR risks associated with specific antibiotics and bacteria in a tropical reservoir.
• To establish a hybrid modeling approach for holistic AMR prediction and risk management.

Main Methods:

• Developed a hybrid modeling framework integrating statistical and process-based models.
• Focused on sulfamethoxazole (SMX) and trimethoprim (TMP) antibiotics, and *Escherichia coli* (*E. coli*) and its resistant variant (EC_SXT).
• Validated the model using withheld data, achieving high performance metrics (NSE > 0.79, ARD < 25%, R² > 0.800).

Main Results:

• Predicted SMX concentrations of 1-15 ng/L and TMP concentrations of 0.5-5 ng/L.
• Predicted *E. coli* abundance from 0 to 5 (log10 MPN/100 mL) and EC_SXT from -1.1 to 3.5 (log10 CFU/100 mL).
• Risk assessment indicated TMP poses a higher risk for AMR development, while SMX presents a greater ecological risk.

Conclusions:

• The study successfully established a validated hybrid modeling framework for predicting AMR in aquatic ecosystems.
• The findings highlight differential risks posed by SMX and TMP, informing targeted management strategies.
• This integrated approach facilitates enhanced risk management frameworks for aquatic environments.