EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Generating realistic artificial human genomes using adversarial autoencoders

Callum Burnard iD, Alban Mancheron, William Ritchie iD

NAR Genomics and Bioinformatics · 2025

DOI: 10.1093/nargab/lqaf101 ↗

Kurz gefasst

Fachartikel in der Zeitschrift NAR Genomics and Bioinformatics, 2025. Verfasst von Callum Burnard und 2 weiteren Autor:innen. Beteiligte Einrichtung: Institut de Génétique Humaine, 34094 Montpellier, France. Hauptthemen laut MEDLINE: Genome, Human, Artificial Intelligence, Genomics. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Agence Nationale de la Recherche.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

A publicly available human genome is both valuable to researchers and a risk for its donor. Many actors could exploit it to extract information about the donor’s health or that of their relatives. Recent efforts have employed artificial intelligence models to simulate genomic data, aiming to create synthetic datasets with scientific merit while preserving patient anonymity. Challenges arise due to the vast amount of data that constitute a complete human genome and the computational resources required. We present a dimension reduction method that combines artificial intelligence with our knowledge of in vivo mutation association mechanisms. This approach enables processing large amounts of data without significant computational resources. Our genome segmentation follows chromosomal recombination hotspots, closely resembling mutation transmission mechanisms. Data from the 1000 Genomes Project are used to train variational autoencoders with a Wasserstein GAN to generate novel data in a two-step process. After optimizing our strategy, our pipeline can generate a simulated population meeting several essential criteria. They are diverse but realistic; the newly generated combinations of mutations follow linkage disequilibrium found in humans. Our pipeline does not reveal the genetic identity of any individual donor, synthesizing genomes that differ from reference samples.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Genome, HumanArtificial IntelligenceGenomicsHumansMutationAlgorithmsLinkage DisequilibriumAutoencoder

Autor:innen und Institutionen

  • Institut de Génétique Humaine, 34094 Montpellier, France.
  • Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier, 34095 Montpellier, France.

ORCID-Profile: Callum Burnard, William Ritchie

Förderung

  • Agence Nationale de la Recherche

Zitieren

APA

Burnard, C., Mancheron, A., & Ritchie, W. (2025). Generating realistic artificial human genomes using adversarial autoencoders. NAR Genomics and Bioinformatics. https://doi.org/10.1093/nargab/lqaf101

Vancouver

Burnard C, Mancheron A, Ritchie W. Generating realistic artificial human genomes using adversarial autoencoders. NAR Genomics and Bioinformatics. 2025. doi:10.1093/nargab/lqaf101

BibTeX
@article{Burnard2025,
  author = {Burnard, Callum and Mancheron, Alban and Ritchie, William},
  title = {Generating realistic artificial human genomes using adversarial autoencoders},
  journal = {NAR Genomics and Bioinformatics},
  year = {2025},
  doi = {10.1093/nargab/lqaf101}
}
Kostenlos starten →