A scalable gene network model of regulatory dynamics in single cells

2025-03-25

AI Summary

This study addresses the challenges of identifying transcriptional regulatory mechanisms from single-cell perturbation data and modeling their dynamics under sparsity, noise, and scalability constraints. We propose scDiff—the first scalable, structured neural differential equation model—designed specifically for single-cell perturbation analysis. scDiff innovatively embeds learnable functional modules into a coupled ordinary differential equation (ODE) framework, integrating gene co-expression network priors with pseudo-temporal sequence modeling, and employs variational inference to enable robust dynamical inference at million-cell scale. Compared to conventional ODE-based approaches, scDiff significantly improves scalability and robustness in temporal regulatory dynamics modeling. On myeloid differentiation and K562 Perturb-seq datasets, it accurately disentangles direct and indirect perturbation effects. In the A549 small-molecule intervention task, scDiff achieves a 32% improvement in single-cell trajectory prediction accuracy over baseline methods.

Technology Category

• Machine Learning: Scalability of ML Systems
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Single-cell data provide high-dimensional measurements of the transcriptional states of cells, but extracting insights into the regulatory functions of genes, particularly identifying transcriptional mechanisms affected by biological perturbations, remains a challenge. Many perturbations induce compensatory cellular responses, making it difficult to distinguish direct from indirect effects on gene regulation. Modeling how gene regulatory functions shape the temporal dynamics of these responses is key to improving our understanding of biological perturbations. Dynamical models based on differential equations offer a principled way to capture transcriptional dynamics, but their application to single-cell data has been hindered by computational constraints, stochasticity, sparsity, and noise. Existing methods either rely on low-dimensional representations or make strong simplifying assumptions, limiting their ability to model transcriptional dynamics at scale. We introduce a Functional and Learnable model of Cell dynamicS, FLeCS, that incorporates gene network structure into coupled differential equations to model gene regulatory functions. Given (pseudo)time-series single-cell data, FLeCS accurately infers cell dynamics at scale, provides improved functional insights into transcriptional mechanisms perturbed by gene knockouts, both in myeloid differentiation and K562 Perturb-seq experiments, and simulates single-cell trajectories of A549 cells following small-molecule perturbations.

Problem

• Modeling gene regulatory dynamics in single-cell data
• Distinguishing direct vs indirect effects of perturbations
• Overcoming computational constraints in dynamical modeling

Innovation

• Scalable gene network model for single cells
• Coupled differential equations for gene regulation
• Accurate inference of cell dynamics from data