Quantitative Structure-Activity Relationship Modeling of the Amplex Ultrared Assay to Predict Thyroperoxidase Inhibitory Activity

Abstract

The thyroid system plays a major role in the regulation of several physiological processes. The dysregulation of the thyroid system caused by the interference of xenobiotics and contaminants may bring to pathologies like hyper- and hypothyroidism and it has been recently correlated with adverse outcomes leading to cancer, obesity, diabetes and neurodevelopmental disorders. Thyroid disruption can occur at several levels. For example, the inhibition of thyroperoxidase (TPO) enzyme, which catalyses the synthesis of thyroid hormones, may cause dysfunctions related to hypothyroidism. The inhibition of the TPO enzyme can occur as a consequence of prolonged exposure to chemical compounds, for this reason it is of utmost importance to identify alternative methods to evaluate the large amount of pollutants and other chemicals that may pose a potential hazard to the human health. In this work, quantitative structure-activity relationship (QSAR) models to predict the TPO inhibitory potential of chemicals are presented. Models are developed by means of several machine learning and data selection approaches, and are based on data obtained in vitro with the Amplex UltraRed-thyroperoxidase (AUR-TPO) assay. Balancing methods and feature selection are applied during model development. Models are rigorously evaluated through internal and external validation. Based on validation results, two models based on Balanced Random Forest (BRF) and K-Nearest Neighbours (KNN) algorithms were selected for a further validation phase, that leads predictive performance (BA = 0.76–0.78 on external data) that is comparable with the reported experimental variability of the AUR-TPO assay (BA ∼0.70). Finally, a consensus between the two models was proposed (BA = 0.82). Based on the predictive performance, these models can be considered suitable for toxicity screening of environmental chemicals.

Prediction of the Neurotoxic Potential of Chemicals Based on Modelling of Molecular Initiating Events Upstream of the Adverse Outcome Pathways of (Developmental) Neurotoxicity

“…As no data were found for TPO inhibition from ChEMBL, the QSAR model for predicting TPO inhibition proposed by Gadaleta et al [ 68 ] was used. The model was developed from data related to the Amplex UltraRed-thyroperoxidase (AUR-TPO) assay.…”

“…The QSAR was based on a BRF developed with the imbalance-learner and scikit-learn Python libraries [ 74 ] and based on 160 DRAGON descriptors [ 75 ] with a training set of 723 chemicals. Additional details on the predictive performance of the model can be found in [ 68 ].…”

“…The MIEs for the 69 chemicals with data for neuroactivity were predicted with the 15 BRFs developed from the entire datasets in Table 4 using the BRF model to predict AUR-TPO from Gadaleta et al [ 68 ], and were profiled with the SMARTS for electrophilic activity compiled by Enoch et al [ 69 ].…”

“…No data found in ChEMBL, QSAR from Gadaleta et al[68] was used 2. Replaced with the use of reactivity SMARTS[69] 3.…”