Single-cell perturbation prediction near experimental reproducibility with GENESIS

Abstract

Single-cell perturbation atlases now measure cellular responses across millions of cells, but the number of possible combinations of biological contexts, perturbations, doses and other experimental variables make exhaustive measurements infeasible. Here, we present GENESIS (Generative ENErgy-based Single-cell Intervention Simulator), a generative framework that predicts single-cell response distributions directly from the covariates defining an experimental condition. GENESIS uses a lightweight neural network trained with the energy score, a strictly proper scoring rule for distributions. We introduce an experiment-calibrated evaluation framework in which prediction error is interpreted relative to the reproducibility of repeated measurements of the same condition. Across the Tahoe-100M small-molecule atlas and a large PBMC cytokine atlas, GENESIS significantly outperforms existing generative models and statistical baselines and approaches experimental reproducibility on independently repeated conditions. Importantly, GENESIS remains accurate when trained on only a fraction of the available experimental conditions or cells, and generated profiles recover experimentally observed context-dependent transcriptional programs. At comparable training-cell budgets, GENESIS achieves more accurate predictions when measurements span more experimental conditions rather than profiling fewer conditions more deeply. Together, these findings suggest a complementary strategy for perturbation studies: broader, lower-depth measurements can expand experimental coverage, while GENESIS augments sparsely sampled populations and predicts responses in unmeasured conditions.

PerturbBridge: Conditional Latent Schr\"{o}dinger Bridge for Single-Cell Perturbation Prediction
Z. Liu, C. Jiang, C. Yang +1
bioRxiv24 Sep 2026PreprintAll rights reservedFull textBioinformatics & computational biologybioRxiv
scDisentangle: sequential disentanglement for single-cell perturbation response prediction
S. El Mejjad, S. Enezari, A. Bedraoui +4
bioRxiv4 Nov 2026PreprintCC BYFull textBioinformatics & computational biologybioRxiv
A split-control criterion for few-shot adaptation in single-cell perturbation prediction
W.-L. Liang, Y.-B. Yang, Q.-Y. Zhang +12
bioRxiv5 Oct 2026PreprintCC BY-NC-NDFull textBioinformatics & computational biologybioRxiv
PopPert: Population-level Joint-Distribution Modeling for Single-Cell Perturbation Prediction
Handong Wang, Jiaxin Qi, Haochen Feng +1
arXiv1 Sep 2026PreprintCC BY-NC-SA 4.0Full textGenetics & genomicsarXiv
Generalizable single-cell perturbation response prediction using energy-guided flow matching
Jianan Wei, Jiajun Hong, Guikun Chen +3
arXiv1 Oct 2026PreprintarXiv licenceFull textCell biology Genetics & genomicsarXiv

Cite

BibTeX RIS

BibTeX

@misc{conci2026singlecell,
  title = {{Single-cell perturbation prediction near experimental reproducibility with GENESIS}},
  author = {R. Conci and D. D'Ascenzo and S. Cultrera di Montesano and S. Raghavan and P. S. Winter and E. M. Van Allen and L. Crawford and W. Lotter},
  howpublished = {bioRxiv},
  year = {2026},
  month = oct,
  doi = {10.64898/2026.10.30.763081},
  url = {https://www.biorxiv.org/content/10.64898/2026.10.30.763081v1},
  note = {CC BY},
}

RIS

TY  - UNPB
TI  - Single-cell perturbation prediction near experimental reproducibility with GENESIS
AU  - R. Conci
AU  - D. D'Ascenzo
AU  - S. Cultrera di Montesano
AU  - S. Raghavan
AU  - P. S. Winter
AU  - E. M. Van Allen
AU  - L. Crawford
AU  - W. Lotter
PY  - 2026
DA  - 2026/10/30
PB  - bioRxiv
DO  - 10.64898/2026.10.30.763081
UR  - https://www.biorxiv.org/content/10.64898/2026.10.30.763081v1
AB  - Single-cell perturbation atlases now measure cellular responses across millions of cells, but the number of possible combinations of biological contexts, perturbations, doses and other experimental variables make exhaustive measurements infeasible. Here, we present GENESIS (Generative ENErgy-based Single-cell Intervention Simulator), a generative framework that predicts single-cell response distributions directly from the covariates defining an experimental condition. GENESIS uses a lightweight neural network trained with the energy score, a strictly proper scoring rule for distributions. We introduce an experiment-calibrated evaluation framework in which prediction error is interpreted relative to the reproducibility of repeated measurements of the same condition. Across the Tahoe-100M small-molecule atlas and a large PBMC cytokine atlas, GENESIS significantly outperforms existing generative models and statistical baselines and approaches experimental reproducibility on independently repeated conditions. Importantly, GENESIS remains accurate when trained on only a fraction of the available experimental conditions or cells, and generated profiles recover experimentally observed context-dependent transcriptional programs. At comparable training-cell budgets, GENESIS achieves more accurate predictions when measurements span more experimental conditions rather than profiling fewer conditions more deeply. Together, these findings suggest a complementary strategy for perturbation studies: broader, lower-depth measurements can expand experimental coverage, while GENESIS augments sparsely sampled populations and predicts responses in unmeasured conditions.
ER  -

Abstract

Related

Cite