KGBS: Knowledge Graph-Bayesian Network for Suspect Screening of Phthalates

A knowledge graph-Bayesian network-driven suspect screening (KGBS) strategy integrates text mining, probabilistic reasoning, and high-resolution mass spectrometry. The framework extracted 54,838 associations between phthalate esters and their metabolites from 3,167 publications to construct a Bayesian inference-embedded knowledge graph. Applied to paired indoor dust and human urine samples, the KGBS platform identified 68 PAEs and 49 metabolites, including 18 PAEs and 14 metabolites newly annotated in the study.

Use Cases

• Prioritize suspect compounds for environmental monitoring based on the Bayesian inference-embedded knowledge graph.
• Reconstruct biotransformation networks of contaminants based on the 54,838 extracted metabolite associations.
• Annotate unknown peaks in HRMS data based on the mechanistically informed spectral library.
• Assess human exposure risk to phthalates based on paired indoor dust and urine sample analysis.

Strengths

• Integrates a large knowledge base of 54,838 associations extracted from 3,167 scientific publications.
• Validated framework identified 68 phthalates and 49 metabolites in real-world samples.
• Provides mechanistically informed support via prioritization scores and diagnostic MS2 fragmentation fingerprints.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect publication bias inherent to the text mining source.

Provenance

Source

Author Dian Wang via figshare.

Collection Method

Associations extracted via text mining from publications, coupled with HRMS workflow.

Freshness

Last updated 2026-04-24 10:44:13