B1252
Title: Zero-inflated Bayesian hierarchical mixture model to address the missing data and dropouts for scRNA-Seq data
Authors: Xiaoping Su - MD Anderson Cancer Center (United States) [presenting]
Abstract: Single-cell RNA-Seq (scRNA-seq) is the most widely used to measure genome-wide gene expression at the single-cell level. One challenge to analyze scRNA-seq is that the majority of expression levels are zeros, which could be either biologically driven (genes not expressing RNA) or technically driven (genes expressing RNA) but not at a sufficient level to be detected by sequencing technology. Another challenge is that the proportion of genes with zero expression level varies substantially across single cells. A zero-inflated Bayesian hierarchical mixture model is proposed to address these challenges. Hierarchical structure is used to account for variation across cells, and a mixture model is used to reflect the two sources of zero expression levels. A simulation study shows that the proposed approach yields accurate estimates.
