Screening of persistent organic pollutants by QSPR classification models: a comparative study

Area of Science:

• Environmental Chemistry
• Computational Chemistry
• Toxicology

Background:

• Environmental persistence is a key factor in assessing the risk of organic compounds.
• Predicting persistence is crucial for identifying potential persistent organic pollutants (POPs).

Purpose of the Study:

• To develop and validate Quantitative Structure-Property Relationship (QSPR) models for predicting the environmental persistence of organic compounds.
• To classify compounds into environmental persistence categories using molecular descriptors.

Main Methods:

• Hierarchical Cluster Analysis (HCA) to define persistence classes based on half-life data.
• Development of QSPR classification models using k-Nearest Neighbors (k-NN), Classification and Regression Trees (CART), and Counterpropagation Artificial Neural Networks (CP-ANN).
• Utilized three interpretable theoretical molecular descriptors for model development.
• External validation using statistical splitting and new datasets, including literature data.

Main Results:

• Successfully developed robust QSPR classification models for predicting environmental persistence.
• Models demonstrated good performance and were validated on new, unseen data.
• Analysis revealed a slight tendency towards overestimation of persistence, aligning with the precautionary principle.

Conclusions:

• The developed structure-based QSPR models are reliable tools for predicting the environmental persistence of diverse organic compounds.
• These models can support the identification and prioritization of potential POPs and facilitate pre-synthesis screening to prevent environmental release.