Deep convolutional and conditional neural networks for large-scale genomic data generation
PLOS Computational Biology · 30.10.2023 · 19 (10) · e1011584
DOI: 10.1371/journal.pcbi.1011584 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS Computational Biology, online erschienen am 30.10.2023. Verfasst von Burak Yelmen und 6 weiteren Autor:innen. Beteiligte Einrichtung: Université Paris-Saclay, CNRS, INRIA, LISN, Paris, France. Hauptthemen laut MEDLINE: Genomics, Learning. Die Arbeit wurde bisher 31-mal zitiert (Crossref) und nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Agence Nationale de la Recherche, Comunidad de Madrid, Banco Santander and the UCM.
Abstract (Originalsprache)
Applications of generative models for genomic data have gained significant momentum in the past few years, with scopes ranging from data characterization to generation of genomic segments and functional sequences. In our previous study, we demonstrated that generative adversarial networks (GANs) and restricted Boltzmann machines (RBMs) can be used to create novel high-quality artificial genomes (AGs) which can preserve the complex characteristics of real genomes such as population structure, linkage disequilibrium and selection signals. However, a major drawback of these models is scalability, since the large feature space of genome-wide data increases computational complexity vastly. To address this issue, we implemented a novel convolutional Wasserstein GAN (WGAN) model along with a novel conditional RBM (CRBM) framework for generating AGs with high SNP number. These networks implicitly learn the varying landscape of haplotypic structure in order to capture complex correlation patterns along the genome and generate a wide diversity of plausible haplotypes. We performed comparative analyses to assess both the quality of these generated haplotypes and the amount of possible privacy leakage from the training data. As the importance of genetic privacy becomes more prevalent, the need for effective privacy protection measures for genomic data increases. We used generative neural networks to create large artificial genome segments which possess many characteristics of real genomes without substantial privacy leakage from the training dataset. In the near future, with further improvements in haplotype quality and privacy preservation, large-scale artificial genome databases can be assembled to provide easily accessible surrogates of real databases, allowing researchers to conduct studies with diverse genomic data within a safe ethical framework in terms of donor privacy.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Université Paris-Saclay, CNRS, INRIA, LISN, Paris, France.
- University of Tartu, Institute of Genomics, Tartu, Estonia.
- Universidad Complutense de Madrid, Departamento de Física Teórica, Madrid, Spain.
- Université d'Évry Val-d'Essonne, Évry-Courcouronnes, France.
ORCID-Profile: Burak Yelmen, Aurélien Decelle
Förderung
- Agence Nationale de la Recherche
- Comunidad de Madrid
- Banco Santander and the UCM
- Fondo Europeo de Desarrollo Regional
Zitieren
APA
Yelmen, B., Decelle, A., Boulos, L. L., Szatkownik, A., Furtlehner, C., Charpiat, G., & Jay, F. (2023). Deep convolutional and conditional neural networks for large-scale genomic data generation. PLOS Computational Biology, 19(10), e1011584. https://doi.org/10.1371/journal.pcbi.1011584
Vancouver
Yelmen B, Decelle A, Boulos LL, Szatkownik A, Furtlehner C, Charpiat G, et al. Deep convolutional and conditional neural networks for large-scale genomic data generation. PLOS Computational Biology. 2023;19(10):e1011584. doi:10.1371/journal.pcbi.1011584
BibTeX
@article{Yelmen2023,
author = {Yelmen, Burak and Decelle, Aurélien and Boulos, Leila Lea and Szatkownik, Antoine and Furtlehner, Cyril and Charpiat, Guillaume and Jay, Flora},
title = {Deep convolutional and conditional neural networks for large-scale genomic data generation},
journal = {PLOS Computational Biology},
year = {2023},
volume = {19},
number = {10},
pages = {e1011584},
doi = {10.1371/journal.pcbi.1011584}
}
EUVIMED