EUVIMEDEUVIMED entdecken
ÜbersichtsarbeitOpen Access

Curated Data In — Trustworthy In Silico Models Out: The Impact of Data Quality on the Reliability of Artificial Intelligence Models as Alternatives to Animal Testing

Vinicius M Alves iD, Scott S Auerbach, Nicole Kleinstreuer, John P Rooney, Eugene N Muratov, Ivan Rusyn, Alexander Tropsha, Charles Schmitt

Alternatives to Laboratory Animals · 2021-05 · 49 (3) · 73-82

DOI: 10.1177/02611929211029635 ↗

Kurz gefasst

Übersichtsarbeit in der Zeitschrift Alternatives to Laboratory Animals, online erschienen am 07.07.2021. Verfasst von Vinicius M Alves und 7 weiteren Autor:innen. Beteiligte Einrichtung: Office of Data Science, Division of the National Toxicology Program (DNTP), 6857National Institute of Environmental Health Sciences (NIEHS), Durham, …. Hauptthemen laut MEDLINE: Animal Testing Alternatives, Artificial Intelligence. Die Arbeit wurde bisher 70-mal zitiert (Crossref) und nennt 86 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIEHS NIH HHS, National Institute of Environmental Health Sciences.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

New Approach Methodologies (NAMs) that employ artificial intelligence (AI) for predicting adverse effects of chemicals have generated optimistic expectations as alternatives to animal testing. However, the major underappreciated challenge in developing robust and predictive AI models is the impact of the quality of the input data on the model accuracy. Indeed, poor data reproducibility and quality have been frequently cited as factors contributing to the crisis in biomedical research, as well as similar shortcomings in the fields of toxicology and chemistry. In this article, we review the most recent efforts to improve confidence in the robustness of toxicological data and investigate the impact that data curation has on the confidence in model predictions. We also present two case studies demonstrating the effect of data curation on the performance of AI models for predicting skin sensitisation and skin irritation. We show that, whereas models generated with uncurated data had a 7–24% higher correct classification rate (CCR), the perceived performance was, in fact, inflated owing to the high number of duplicates in the training set. We assert that data curation is a critical step in building computational models, to help ensure that reliable predictions of chemical toxicity are achieved through use of the models.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Animal Testing AlternativesArtificial IntelligenceAnimalsComputer SimulationData AccuracyReproducibility of Results

Stichwörter der Autor:innen

QSARartificial intelligencedata curationdata qualitydata reproducibility

Autor:innen und Institutionen

  • Office of Data Science, Division of the National Toxicology Program (DNTP), 6857National Institute of Environmental Health Sciences (NIEHS), Durham, NC, USA.
  • Toxinformatics Group, Predictive Toxicology Branch, DNTP, NIEHS, Durham, NC, USA.
  • National Toxicology Program Interagency Center for the Evaluation of Alternative Toxicological Methods, Scientific Director's Office, DNTP, NIEHS, Durham, NC, USA.
  • Integrated Laboratory Systems, LLC, Morrisville, NC, USA.
  • Laboratory for Molecular Modeling, UNC Eshelman School of Pharmacy, 2331The University of North Carolina at Chapel Hill, NC, USA.
  • Department of Pharmaceutical Sciences, Federal University of Paraiba, Joao Pessoa, Paraiba, Brazil.

ORCID-Profile: Vinicius M Alves

Förderung

  • NIEHS NIH HHS
  • National Institute of Environmental Health Sciences

Zitieren

APA

Alves, V. M., Auerbach, S. S., Kleinstreuer, N., Rooney, J. P., Muratov, E. N., Rusyn, I., Tropsha, A., & Schmitt, C. (2021). Curated Data In — Trustworthy In Silico Models Out: The Impact of Data Quality on the Reliability of Artificial Intelligence Models as Alternatives to Animal Testing. Alternatives to Laboratory Animals, 49(3), 73-82. https://doi.org/10.1177/02611929211029635

Vancouver

Alves VM, Auerbach SS, Kleinstreuer N, Rooney JP, Muratov EN, Rusyn I, et al. Curated Data In — Trustworthy In Silico Models Out: The Impact of Data Quality on the Reliability of Artificial Intelligence Models as Alternatives to Animal Testing. Alternatives to Laboratory Animals. 2021;49(3):73-82. doi:10.1177/02611929211029635

BibTeX
@article{Alves2021,
  author = {Alves, Vinicius M and Auerbach, Scott S and Kleinstreuer, Nicole and Rooney, John P and Muratov, Eugene N and Rusyn, Ivan and Tropsha, Alexander and Schmitt, Charles},
  title = {Curated Data In — Trustworthy In Silico Models Out: The Impact of Data Quality on the Reliability of Artificial Intelligence Models as Alternatives to Animal Testing},
  journal = {Alternatives to Laboratory Animals},
  year = {2021},
  volume = {49},
  number = {3},
  pages = {73--82},
  doi = {10.1177/02611929211029635}
}
Kostenlos starten →