文章基本信息

标题：Statistical methods for analysis of single-cell RNA-sequencing data
本地全文：下载
作者：Samarendra Das ; Shesh N. Rai
期刊名称：MethodsX
印刷版ISSN：2215-0161
电子版ISSN：2215-0161
出版年度：2021
卷号：8
页码：1-18
DOI：10.1016/j.mex.2021.101580
语种：English
出版社：Elsevier
摘要：AbstractSingle-cell RNA-sequencing (scRNA-seq) is a recent high-throughput genomic technology used to study the expression dynamics of genes at single-cell level. Analyzing the scRNA-seq data in presence of biological confounding factors including dropout events is a challenging task. Thus, this article presents a novel statistical approach for various analyses of the scRNA-seq Unique Molecular Identifier (UMI) counts data. The various analyses include modeling and fitting of observed UMI data, cell type detection, estimation of cell capture rates, estimation of gene specific model parameters, estimation of the sample mean and sample variance of the genes,etc. Besides, the developed approach is able to perform differential expression, and other downstream analyses that consider the molecular capture process in scRNA-seq data modeling. Here, the external spike-ins data can also be used in the approach for better results. The unique feature of the method is that it considers the biological process that leads to severe dropout events in modeling the observed UMI counts of genes.• The differential expression analysis of observed scRNA-seq UMI counts data is performed after adjustment for cell capture rates.• The statistical approach performs downstream differential zero inflation analysis, classification of influential genes, and selection of top marker genes.• Cell auxiliaries including cell clusters and other cell variables (e.g., cell cycle, cell phase) are used to remove unwanted variation to perform statistical tests reliably.Graphical abstractDisplay Omitted
关键词：Zero inflated negative binomial model;Molecular capture model;Observed UMI count;True UMI count;Mean;Zero Inflation;Overdispersion