EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Building species trait-specific nano-QSARs: Model stacking, navigating model uncertainties and limitations, and the effect of dataset size

Surendra Balraadjsing, Willie J G M Peijnenburg, Martina G Vijver

Environment International · 2024-06 · 188 · 108764

DOI: 10.1016/j.envint.2024.108764 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Environment International, 2024-06. Verfasst von Surendra Balraadjsing und 2 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Environmental Sciences (CML), Leiden University, PO Box 9518, 2300 RA Leiden, the Netherlands. Electronic address: s.balraadjsing@cml.le…. Hauptthemen laut MEDLINE: Quantitative Structure-Activity Relationship. Die Arbeit wurde bisher 23-mal zitiert (Crossref) und nennt 77 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Commission, European Research Council.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

A strong need exists for broadly applicable nano-QSARs, capable of predicting toxicological outcomes towards untested species and nanomaterials, under different environmental conditions. Existing nano-QSARs are generally limited to only a few species but the inclusion of species characteristics into models can aid in making them applicable to multiple species, even when toxicity data is not available for biological species. Species traits were used to create classification- and regression machine learning models to predict acute toxicity towards aquatic species for metallic nanomaterials. Afterwards, the individual classification- and regression models were stacked into a meta-model to improve performance. Additionally, the uncertainty and limitations of the models were assessed in detail (beyond the OECD principles) and it was investigated whether models would benefit from the addition of more data. Results showed a significant improvement in model performance following model stacking. Investigation of model uncertainties and limitations highlighted the discrepancy between the applicability domain and accuracy of predictions. Data points outside of the assessed chemical space did not have higher likelihoods of generating inadequate predictions or vice versa. It is therefore concluded that the applicability domain does not give complete insight into the uncertainty of predictions and instead the generation of prediction intervals can help in this regard. Furthermore, results indicated that an increase of the dataset size did not improve model performance. This implies that larger dataset sizes may not necessarily improve model performance while in turn also meaning that large datasets are not necessarily required for prediction of acute toxicity with nano-QSARs.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Quantitative Structure-Activity RelationshipUncertaintyNanostructuresAnimalsMachine LearningAquatic Organisms

Stichwörter der Autor:innen

Machine learningMetallic nanoparticlesOECD principlesPrediction uncertaintyQSARSpecies traits

Autor:innen und Institutionen

  • Institute of Environmental Sciences (CML), Leiden University, PO Box 9518, 2300 RA Leiden, the Netherlands. Electronic address: s.balraadjsing@cml.leidenuniv.nl.
  • Institute of Environmental Sciences (CML), Leiden University, PO Box 9518, 2300 RA Leiden, the Netherlands; Centre for Safety of Substances and Products, National Institute of Public Health and the Environment (RIVM), P…
  • Institute of Environmental Sciences (CML), Leiden University, PO Box 9518, 2300 RA Leiden, the Netherlands.

Förderung

  • European Commission
  • European Research Council

Zitieren

APA

Balraadjsing, S., J G M Peijnenburg, W., & Vijver, M. G. (2024). Building species trait-specific nano-QSARs: Model stacking, navigating model uncertainties and limitations, and the effect of dataset size. Environment International, 188, 108764. https://doi.org/10.1016/j.envint.2024.108764

Vancouver

Balraadjsing S, J G M Peijnenburg W, Vijver MG. Building species trait-specific nano-QSARs: Model stacking, navigating model uncertainties and limitations, and the effect of dataset size. Environment International. 2024;188:108764. doi:10.1016/j.envint.2024.108764

BibTeX
@article{Balraadjsing2024,
  author = {Balraadjsing, Surendra and J G M Peijnenburg, Willie and Vijver, Martina G},
  title = {Building species trait-specific nano-QSARs: Model stacking, navigating model uncertainties and limitations, and the effect of dataset size},
  journal = {Environment International},
  year = {2024},
  volume = {188},
  pages = {108764},
  doi = {10.1016/j.envint.2024.108764}
}
Kostenlos starten →