Deep learning-based virtual perturbation reveals increasing heterogeneity of gene regulatory responses with…

Abstract

A central hypothesis in aging biology proposes that the gradual accumulation of stochastic molecular damage drives transcriptional deregulation (TD), ultimately leading to functional decline. However, existing single-cell analyses largely rely on static expression measurements and provide limited resolution of how regulatory responses change at the level of individual cells. Here, we leverage DKO, a deep learning framework that predicts cell-specific virtual gene perturbation effects from unperturbed transcriptomes, capturing cell-to-cell differences in regulatory context without requiring a reconstructed gene regulatory network or matched perturbation data. Applying DKO to single-cell transcriptomic datasets from humans, mice, Drosophila, and C. elegans revealed that the transcriptional consequences of perturbing the same gene became increasingly heterogeneous across cells with age, indicating that gene regulatory responses become less consistent and more dependent on cellular context. Furthermore, we find that cells with higher somatic mutation burdens have greater knockout-response variability, linking accumulated molecular damage to regulatory inconsistency. Within the same cellular populations, higher impact variability was also associated with higher senescence scores, linking transcriptional dysregulation to senescence-associated states. Our findings identify a progressive loss of regulatory robustness at both the gene and cell levels as a conserved feature of aging, linking stochastic molecular damage to increasing heterogeneity in cellular regulatory responses.

SPECTRA: predicting cellular perturbation responses with Graph Learning over Gene Regulatory Networks
M. Calabro, P. Sheehan, F. Cambuli +1
bioRxiv21 Sep 2026PreprintCC BYFull textSystems & synthetic biologybioRxiv
Where single-cell foundation models help: a benchmark of annotation, perturbation and regulatory inference against simple baselines
F. A. Cupaioli, I. Merelli
bioRxiv3 Nov 2026PreprintAll rights reservedFull textBioinformatics & computational biologybioRxiv
Mitochondrial perturbation adapts the proteome from early to advanced immune responses
J. Alvarez, S. Rivera, S. Roberts +6
bioRxiv7 Sep 2026PreprintAll rights reservedFull textImmunologybioRxiv
DELPHAI predicts heterogeneous perturbation responses with learned single-cell fitness
X. Zhang, H. Wu, H. Liu
bioRxiv1 Jul 2026PreprintCC BY-NCFull textBioinformatics & computational biologybioRxiv
Physiological Responses of European Seabass (Dicentrarchus labrax) Juveniles to Long-Term Feeding with Increasing Levels of EPA and DHA: Effects of Thermal Challenge.
Lúcia Vieira, Diogo Amaral, Bruno Silveira +9
Animals : an open access journal from MDPI4 Oct 2026Peer-reviewedCC BYFull textEcologyEurope PMC

Cite

BibTeX RIS

BibTeX

@misc{he2026deep,
  title = {{Deep learning-based virtual perturbation reveals increasing heterogeneity of gene regulatory responses with aging}},
  author = {Yongjian He and Julia Hanna Haddad Guez and Xu-Wen Wang and Orr Levy},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.19029},
  eprint = {2610.19029},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.19029},
}

RIS

TY  - UNPB
TI  - Deep learning-based virtual perturbation reveals increasing heterogeneity of gene regulatory responses with aging
AU  - Yongjian He
AU  - Julia Hanna Haddad Guez
AU  - Xu-Wen Wang
AU  - Orr Levy
PY  - 2026
DA  - 2026/10/13
PB  - arXiv
DO  - 10.48550/arxiv.2610.19029
UR  - https://arxiv.org/abs/2610.19029
AB  - A central hypothesis in aging biology proposes that the gradual accumulation of stochastic molecular damage drives transcriptional deregulation (TD), ultimately leading to functional decline. However, existing single-cell analyses largely rely on static expression measurements and provide limited resolution of how regulatory responses change at the level of individual cells. Here, we leverage DKO, a deep learning framework that predicts cell-specific virtual gene perturbation effects from unperturbed transcriptomes, capturing cell-to-cell differences in regulatory context without requiring a reconstructed gene regulatory network or matched perturbation data. Applying DKO to single-cell transcriptomic datasets from humans, mice, Drosophila, and C. elegans revealed that the transcriptional consequences of perturbing the same gene became increasingly heterogeneous across cells with age, indicating that gene regulatory responses become less consistent and more dependent on cellular context. Furthermore, we find that cells with higher somatic mutation burdens have greater knockout-response variability, linking accumulated molecular damage to regulatory inconsistency. Within the same cellular populations, higher impact variability was also associated with higher senescence scores, linking transcriptional dysregulation to senescence-associated states. Our findings identify a progressive loss of regulatory robustness at both the gene and cell levels as a conserved feature of aging, linking stochastic molecular damage to increasing heterogeneity in cellular regulatory responses.
ER  -

Abstract

Related

Cite