GenoME: a MoE-based generative model for individualized, multimodal prediction and perturbation of genomic profiles
Nucleic Acids Research · 2026
DOI: 10.1093/nar/gkag902 ↗Kurz gefasst
Fachartikel in der Zeitschrift Nucleic Acids Research, 2026. Verfasst von Jiachen Wei und 3 weiteren Autor:innen. Beteiligte Einrichtung: Changping Laboratory, Beijing 102206, China. Hauptthemen laut MEDLINE: Genomics, Genome, Models, Genetic. Die Arbeit nennt 62 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: National Science and Technology, National Natural Science Foundation of China, New Cornerstone Science Foundation.
Abstract (Originalsprache)
The non-coding genome operates through a complex, multiscale regulatory system where regulated gene expressions are closely associated with cell-type-specific histone modifications, transcription factor binding, and 3D conformation. Developing computational models that can integrate these patterns to predict and interpret the regulatory system remains challenging. Here, we present GenoME, a Mixture of Experts (MoE)-based generative model that uses DNA sequence and cell-type-specific ATAC-seq signals to predict a unified genomic profiles encompassing epigenomics, transcriptomics, and chromatin architecture at base-pair to kilobase resolutions. GenoME enables multiscale predictions for held-out genomic regions and, critically, generalizes to predict the full regulatory landscape of unseen or individualized cell types from a single ATAC-seq input. We equip GenoME with an in silico perturbation framework that accurately forecasts the multimodal consequences of genetic perturbations and identifies functional enhancer–promoter connections, outperforming specialized models like Activity-by-Contact. These predictions can also be used to decipher the transcription factor grammar of cell-type-specific enhancers. GenoME thus provides a versatile, all-in-one platform for generative modeling, cross-cell-type generalization, and causal mechanistic investigation of the multiscale regulatory genome.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Changping Laboratory, Beijing 102206, China.
- School of Science, Westlake University, Hangzhou 310030, China.
- Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871, China.
- Biomedical Pioneering Innovation Center (BIOPIC), Peking University, Beijing 100871, China.
ORCID-Profile: Jiachen Wei, Yi Qin Gao
Förderung
- National Science and Technology
- National Natural Science Foundation of China
- New Cornerstone Science Foundation
Zitieren
APA
Wei, J., Xue, Y., Chai, H., & Gao, Y. Q. (2026). GenoME: a MoE-based generative model for individualized, multimodal prediction and perturbation of genomic profiles. Nucleic Acids Research. https://doi.org/10.1093/nar/gkag902
Vancouver
Wei J, Xue Y, Chai H, Gao YQ. GenoME: a MoE-based generative model for individualized, multimodal prediction and perturbation of genomic profiles. Nucleic Acids Research. 2026. doi:10.1093/nar/gkag902
BibTeX
@article{Wei2026,
author = {Wei, Jiachen and Xue, Yue and Chai, Hao and Gao, Yi Qin},
title = {GenoME: a MoE-based generative model for individualized, multimodal prediction and perturbation of genomic profiles},
journal = {Nucleic Acids Research},
year = {2026},
doi = {10.1093/nar/gkag902}
}
EUVIMED