Latent Causal Diffusions for Single-Cell Perturbation Modeling

TL;DR

This paper introduces the latent causal diffusion (LCD) model for single-cell gene expression, which predicts responses to perturbations and uncovers causal gene regulatory networks by combining generative modeling with causal inference.

Contribution

The paper presents LCD, a novel generative model for single-cell transcriptomics that predicts perturbation effects and learns causal gene regulation dynamics, along with CLIPR for causal effect estimation.

Findings

LCD outperforms existing methods in predicting unseen perturbation responses.

CLIPR accurately recovers causal gene relationships in simulated and real data.

The framework identifies functional gene modules and causal structures beyond differential expression analysis.

Abstract

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.