Toward NGRA for safer phthalate substitution: An AOP-informed IATA case study integrating AI-based predictions and in vitro NAM evidence for DART prioritization
- 1School of Environmental Engineering, University of Seoul, 163 Seoulsiripdae-ro, Dongdaemun-gu, Seoul 02504, South Korea
- 2Corresponding author.
Abstract
Plasticizer substitution can address concerns about regulated legacy phthalates, but decisions based on incomplete hazard and exposure evidence may result in regrettable substitution. We evaluated whether an adverse outcome pathway (AOP)-informed framework integrating predictions from ToxBAI, an AI-based toxicity prediction platform, with experimental new approach methodologies (NAMs) evidence could support screening-level prioritization of plasticizers for male reproductive developmental toxicity within the broader domain of developmental and reproductive toxicity (DART). Four legacy phthalates and four alternative plasticizers were assessed using regulatory data, literature-derived in vitro evidence, ToxCast bioactivity data, ToxBAI toxicity predictions, in vitro-to-in vivo extrapolation-derived human-equivalent doses (HEDs), and exposure records from the Comparative Toxicogenomics Database. Evidence was mapped onto an AOP network involving steroidogenesis, androgen signaling, and male reproductive tract malformations. Di(2-ethylhexyl) phthalate and dibutyl phthalate showed coherent evidence across regulatory outcomes, AOP-linked key events, and HED-scaled bioactivity, supporting high concern for DART. Diethyl phthalate and di-n-octyl phthalate showed less consistent mechanistic support. Alternative plasticizers generally exhibited fewer active signals, but this pattern should be interpreted given limited assay coverage and sparse concentration–response data for DART-related mechanisms. Internal exposure evidence for sensitive developmental and reproductive windows was sparse. Isolated signals involving PPARα activation, testosterone-related responses, and androgen receptor activity indicated that these alternatives cannot yet be confirmed as safer substitutes. Overall, AOP-informed integration of ToxBAI predictions with existing NAMs, regulatory, and exposure evidence provided a transparent screening-level prioritization framework and identified uncertainties relevant to safer substitution. This hazard-focused IATA case study provides a mechanistic foundation for future exposure-informed NGRA.
Related
- Integrating sequence-to-omics predictions and personal multi-omics for rare variant prioritization
- Dynamic machine learning architecture integrating time-phased predictions for acute kidney injury following extracorporeal membrane oxygenation
- ICONIC: An R Package for Integrating Instrumental Variable- and Negative-Control-Informed Causal Discovery and Diagnostics in Multiomic Studies
- Dose-Informed Surgical Margin in Mandibular Osteoradionecrosis Surgery: Integrating Radiation Biology into Osteotomy Planning.
- Recalibrating anterior cruciate ligament reconstruction rehabilitation: integrating mechanistic evidence, criterion-driven progression, and risk-informed return-to-sport decisions: a narrative review.
BibTeX
@article{kim2026toward,
title = {{Toward NGRA for safer phthalate substitution: An AOP-informed IATA case study integrating AI-based predictions and in vitro NAM evidence for DART prioritization}},
author = {Donghyeon Kim and Chaein Chong and Ga-eul Yang and Hyunwoo Kim and Jinhee Choi},
journal = {Environment International},
year = {2026},
month = oct,
doi = {10.1016/j.envint.2026.110556},
url = {https://doi.org/10.1016/j.envint.2026.110556},
note = {CC BY},
}
RIS
TY - JOUR
TI - Toward NGRA for safer phthalate substitution: An AOP-informed IATA case study integrating AI-based predictions and in vitro NAM evidence for DART prioritization
AU - Donghyeon Kim
AU - Chaein Chong
AU - Ga-eul Yang
AU - Hyunwoo Kim
AU - Jinhee Choi
PY - 2026
DA - 2026/10/30
JO - Environment International
DO - 10.1016/j.envint.2026.110556
UR - https://doi.org/10.1016/j.envint.2026.110556
AB - Plasticizer substitution can address concerns about regulated legacy phthalates, but decisions based on incomplete hazard and exposure evidence may result in regrettable substitution. We evaluated whether an adverse outcome pathway (AOP)-informed framework integrating predictions from ToxBAI, an AI-based toxicity prediction platform, with experimental new approach methodologies (NAMs) evidence could support screening-level prioritization of plasticizers for male reproductive developmental toxicity within the broader domain of developmental and reproductive toxicity (DART). Four legacy phthalates and four alternative plasticizers were assessed using regulatory data, literature-derived in vitro evidence, ToxCast bioactivity data, ToxBAI toxicity predictions, in vitro-to-in vivo extrapolation-derived human-equivalent doses (HEDs), and exposure records from the Comparative Toxicogenomics Database. Evidence was mapped onto an AOP network involving steroidogenesis, androgen signaling, and male reproductive tract malformations. Di(2-ethylhexyl) phthalate and dibutyl phthalate showed coherent evidence across regulatory outcomes, AOP-linked key events, and HED-scaled bioactivity, supporting high concern for DART. Diethyl phthalate and di-n-octyl phthalate showed less consistent mechanistic support. Alternative plasticizers generally exhibited fewer active signals, but this pattern should be interpreted given limited assay coverage and sparse concentration–response data for DART-related mechanisms. Internal exposure evidence for sensitive developmental and reproductive windows was sparse. Isolated signals involving PPARα activation, testosterone-related responses, and androgen receptor activity indicated that these alternatives cannot yet be confirmed as safer substitutes. Overall, AOP-informed integration of ToxBAI predictions with existing NAMs, regulatory, and exposure evidence provided a transparent screening-level prioritization framework and identified uncertainties relevant to safer substitution. This hazard-focused IATA case study provides a mechanistic foundation for future exposure-informed NGRA.
SN - 0160-4120
ER -