Identifying Differentially Expressed Genes of Zero Inflated Single Cell RNA Sequencing Data Using Mixed Model Score Tests.

Zhiqiang He,Yueyun Pan,Hui Wang,Fang Shao

doi:10.3389/fgene.2021.616686

Zhiqiang He, Yueyun Pan + Show 2 more

Open Access

https://doi.org/10.3389/fgene.2021.616686

Copy DOI

Abstract

Single cell RNA sequencing (scRNA-seq) allows quantitative measurement and comparison of gene expression at the resolution of single cells. Ignoring the batch effects and zero inflation of scRNA-seq data, many proposed differentially expressed (DE) methods might generate bias. We propose a method, single cell mixed model score tests (scMMSTs), to efficiently identify DE genes of scRNA-seq data with batch effects using the generalized linear mixed model (GLMM). scMMSTs treat the batch effect as a random effect. For zero inflation, scMMSTs use a weighting strategy to calculate observational weights for counts independently under zero-inflated and zero-truncated distributions. Counts data with calculated weights were subsequently analyzed using weighted GLMMs. The theoretical null distributions of the score statistics were constructed by mixed Chi-square distributions. Intensive simulations and two real datasets were used to compare edgeR-zinbwave, DESeq2-zinbwave, and scMMSTs. Our study demonstrates that scMMSTs, as supplement to standard methods, are advantageous to define DE genes of zero-inflated scRNA-seq data with batch effects.

Highlights

In modern biology, transcriptomics has been widely used to elucidate the molecular basis of biological processes and diseases (Van den Berge et al, 2018)
Bath effects are estimated as random effects under the reduced null models of generalized linear mixed model (GLMM)
More complicated models GLMMs are considered, single cell mixed model score tests (scMMSTs) are computationally affordable compared to other differentially expressed (DE) methods

Summary

Introduction

Transcriptomics has been widely used to elucidate the molecular basis of biological processes and diseases (Van den Berge et al, 2018). ScRNA-seq has been treated as an effective method to study cellular heterogeneity in complex biological systems, and is being applied by more researchers in various biological processes, such as stem cell development and differentiation, embryonic organ development, tumors, immunology, and neurology (Tang et al, 2009; McEvoy et al, 2011; Zeisel et al, 2015; Chu et al, 2016; Papalexi and Satija, 2018; Sun et al, 2019). For bulk RNA-seq and scRNA-seq data, batch effects conventionally were treated as the non-biological differences that occurs when samples or cells are measured in distinct batches. For DE analysis, it was recommended that DE testing should be conducted on measure data with covariates including the batch information in the model design, not on batch corrected data (Luecken and Theis, 2019)

Methods

Results

Conclusion