A deep generative model for single-cell RNA sequencing with application to detecting differentially expressed genes

TL;DR

This paper introduces a scalable deep probabilistic model for single-cell RNA sequencing data that effectively accounts for technical variability and batch effects, outperforming existing methods in differential gene expression analysis.

Contribution

The authors develop a flexible neural network-based probabilistic model with scalable inference for single-cell RNA-seq data, including extensions for batch effects and differential expression testing.

Findings

Model scales to over one million cells

Outperforms ZIFA and ZINB-WaVE on multiple tasks

Provides a Bayesian framework for differential expression

Abstract

We propose a probabilistic model for interpreting gene expression levels that are observed through single-cell RNA sequencing. In the model, each cell has a low-dimensional latent representation. Additional latent variables account for technical effects that may erroneously set some observations of gene expression levels to zero. Conditional distributions are specified by neural networks, giving the proposed model enough flexibility to fit the data well. We use variational inference and stochastic optimization to approximate the posterior distribution. The inference procedure scales to over one million cells, whereas competing algorithms do not. Even for smaller datasets, for several tasks, the proposed procedure outperforms state-of-the-art methods like ZIFA and ZINB-WaVE. We also extend our framework to take into account batch effects and other confounding factors and propose a natural Bayesian hypothesis framework for differential expression that outperforms tradition DESeq2.