Latent causal diffusions for single-cell perturbation modeling
Proceedings of the National Academy of Sciences · 2026
DOI: 10.1073/pnas.2602630123 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the National Academy of Sciences, 2026. Verfasst von Lars Lorch und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, Eidgenössische Technische Hochschule (ETH) Zürich, Zürich 8092, Switzerland. Hauptthemen laut MEDLINE: Single-Cell Analysis, Models, Genetic. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 79 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIDDK/NIH, ONR, NCCIH NIH HHS.
Abstract (Originalsprache)
Perturbation screens hold the potential to systematically map regulatory processes at single-cell resolution, yet modeling and predicting transcriptome-wide responses to perturbations remains a major computational challenge. Existing methods often underperform simple baselines, fail to disentangle measurement noise from biological signal, and provide limited insight into the causal structure governing cellular responses. Here, we present the latent causal diffusion (LCD), a generative model that frames single-cell gene expression as a stationary diffusion process observed under measurement noise. LCD outperforms established approaches in predicting the distributional shifts of unseen perturbation combinations in single-cell RNA-sequencing screens while simultaneously learning a mechanistic dynamical system of gene regulation. To interpret these learned dynamics, we develop an approach we call causal linearization via perturbation responses (CLIPR), which yields an approximation of the direct causal effects between all genes modeled by the diffusion. CLIPR provably identifies causal effects under a linear drift assumption and recovers causal structure in both simulated systems and a genome-wide perturbation screen, where it clusters genes into coherent functional modules and resolves causal relationships that standard differential expression analysis cannot. The LCD-CLIPR framework bridges generative modeling with causal inference to predict unseen perturbation effects and map the underlying regulatory mechanisms of the transcriptome.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Computer Science, Eidgenössische Technische Hochschule (ETH) Zürich, Zürich 8092, Switzerland.
- Laboratory for Information and Decision Systems, Massachusetts Institute of Technology (MIT), Cambridge, MA 02139.
- Eric and Wendy Schmidt Center, Broad Institute of MIT and Harvard, Cambridge, MA 02142.
- School of Computer and Communication Sciences, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne 1015, Switzerland.
- School of Life Sciences, Swiss Institute for Experimental Cancer Research, EPFL, Lausanne 1015, Switzerland.
- Max Planck Institute for Intelligent Systems, Tübingen 72076, Germany.
ORCID-Profile: Lars Lorch, Jiaqi Zhang, Charlotte Bunne, Andreas Krause, Bernhard Schölkopf, Caroline Uhler
Förderung
- NIDDK/NIH
- ONR
- NCCIH NIH HHS
- Eric and Wendy Schmidt Center at the Broad Institute
- NCCIH/NIH
- German Research Foundation
- NCCR Automation
- NIDDK NIH HHS
Zitieren
APA
Lorch, L., Zhang, J., Bunne, C., Krause, A., Schölkopf, B., & Uhler, C. (2026). Latent causal diffusions for single-cell perturbation modeling. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2602630123
Vancouver
Lorch L, Zhang J, Bunne C, Krause A, Schölkopf B, Uhler C. Latent causal diffusions for single-cell perturbation modeling. Proceedings of the National Academy of Sciences. 2026. doi:10.1073/pnas.2602630123
BibTeX
@article{Lorch2026,
author = {Lorch, Lars and Zhang, Jiaqi and Bunne, Charlotte and Krause, Andreas and Schölkopf, Bernhard and Uhler, Caroline},
title = {Latent causal diffusions for single-cell perturbation modeling},
journal = {Proceedings of the National Academy of Sciences},
year = {2026},
doi = {10.1073/pnas.2602630123}
}
EUVIMED