From Data to Mechanism: A Knowledge Graph-Bayesian Network-Driven Suspect Screening (KGBS) Strategy for Phthalate...

Area of Science:

• Environmental Chemistry
• Metabolomics
• Toxicology

Background:

• Phthalate esters (PAEs) are widely used, leading to complex human exposure and metabolic transformations.
• Understanding these biotransformation pathways is crucial for assessing human health risks.

Purpose of the Study:

• To develop a systematic strategy for elucidating PAE biotransformation relationships.
• To identify and characterize PAEs and their metabolites in human samples.

Main Methods:

• Developed a knowledge graph-Bayesian network-driven suspect screening (KGBS) strategy.
• Integrated text mining, probabilistic reasoning, and high-resolution mass spectrometry (HRMS).
• Constructed a Bayesian inference-embedded knowledge graph from 3167 publications, extracting 54,838 PAE-metabolite associations.

Main Results:

• Applied the KGBS framework to indoor dust and human urine samples.
• Identified 68 PAEs and 49 metabolites, with 18 PAEs and 14 metabolites newly annotated.
• Prioritization scores and MS2 fragmentation fingerprints supported metabolic linkages.

Conclusions:

• The KGBS strategy enables pathway-centric interpretation of contaminants and their transformation products.
• Established a scalable, self-evolving strategy for reconstructing biotransformation networks.
• Offers new perspectives for exposome characterization and human exposure risk assessment.