An automated toxicity based prioritization framework for fast chemical characterization in non-targeted analysis

Area of Science:

• Environmental Chemistry
• Analytical Chemistry
• Toxicology

Background:

• Non-targeted analysis (NTA) using liquid chromatography-high-resolution mass spectrometry is crucial for identifying environmental pollutants.
• Prioritizing candidate compounds from large mass spectrometry datasets is a significant challenge.
• Assessing exposure potential requires evaluating both toxicity and identification confidence.

Purpose of the Study:

• To develop an automated workflow for prioritizing environmental chemical candidates identified through NTA.
• To integrate toxicity data and identification evidence for effective candidate ranking.
• To reduce the complexity of suspect lists generated from environmental samples.

Main Methods:

• Development of an R package, "NTAprioritization.R", for automated prioritization.
• Rating candidate identification levels based on spectral matching and retention time prediction.
• Rating candidate toxicity using endpoint data or the ToxPi score.
• Tiered classification of candidates from Tier 1 (highest priority) to Tier 5 (lowest priority).

Main Results:

• The workflow successfully reduced a candidate list of over 6,982 compounds to 2,779.
• Prioritization resulted in 5 distinct tiers, effectively ranking compounds by exposure potential.
• Validation using spiked sludge water samples identified 21 out of 28 target pollutants within the prioritized tiers.
• The method demonstrated efficiency in screening environmental pollutants.

Conclusions:

• The "NTAprioritization.R" package provides a valuable tool for rapid screening of environmental pollutants.
• Automated prioritization based on toxicity and identification evidence enhances the efficiency of NTA workflows.
• This approach aids in focusing resources on the most relevant environmental contaminants.