Scalable Single-Cell Gene Expression Generation with Latent Diffusion Models
Preprint
Abstract
Computational modeling of single-cell gene expression is crucial for understanding cellular processes, but generating realistic expression profiles remains a major challenge. This difficulty arises from the count nature of gene expression data and complex latent dependencies among genes. Existing generative models often impose artificial gene orderings or rely on shallow neural network architectures. We introduce a scalable latent diffusion model for single-cell gene expression data, which we refer to as scLDM, that respects the fundamental exchangeability property of the data. Our VAE uses fixed-size latent variables leveraging a unified Multi-head Cross-Attention Block (MCAB) architecture, which serves dual roles: permutation-invariant pooling in the encoder and permutation-equivariant unpooling in the decoder. We enhance this framework by replacing the Gaussian prior with a latent diffusion model using Diffusion Transformers and linear interpolants, enabling high-quality generation with multi-conditional classifier-free guidance. We show its superior performance in a variety of experiments for both observational and perturbational single-cell data, as well as downstream tasks like cell-level classification.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- Squidiff: predicting cellular development and responses to perturbations using a diffusion modelNature Methods, 2025
- Predicting cell morphological responses to perturbations using generative modelingNature Communications, 2025
- scGPT: toward building a foundation model for single-cell multi-omics using generative AINature Methods, 2024
- Learning single-cell perturbation responses using neural optimal transportNature Methods, 2023
- Deep generative modeling of transcriptional dynamics for RNA velocity analysis in single cellsNature Methods, 2023
- The scverse project provides a computational ecosystem for single-cell omics data analysisNature Biotechnology, 2023
- A Python library for probabilistic analysis of single-cell omics dataNature Biotechnology, 2022
- decoupleR: ensemble of computational methods to infer biological activities from omics dataBioinformatics Advances, 2022
- Realistic in silico generation and augmentation of single-cell RNA-seq data using generative adversarial networksNature Communications, 2020
- Deep generative modeling for single-cell transcriptomicsNature Methods, 2018