Work

Addressing Pseudoreplication in Differential Expression Analysis of Single-cell Rna-sequencing Data

Year: 2021
Type: dissertation
Abstract: Single-cell technologies offer a unique opportunity to deepen our understanding of cellular transcription and cellular interaction. However, single-cell RNA-seq data have a hierarchical correlation structure that must be accounted for during differential expression analysis. Computing differential expression analysis with single-cell data as if each cell were independent leads to biased inference and highly inflated type 1 error. In Chapter II of this dissertation, we demonstrate the rationale for utilizing mixed-effects models with a random effect for individual to explicitly model the hierarchical structure of single-cell RNA-seq data. As the field continues to expand, and more single-cell datasets are generated, the application of mixed-effects models with a random effect for individual to compute differential expression analysis will be critical to improving robustness and reproducibility in single-cell research. In Chapter III, we provide a first-of-its-kind R-package designed to approximate the power for tests of differential expression for binary and continuous phenotypes based on user-specified number of independent experimental units (e.g., individuals) and cells within the experimental unit. Specifically, this software simulates gene dropout rates, intra-individual dispersion, inter-individual variation, variable or fixed number of cells per individual, and the correlation among cells within an individual. This R-package is an important addition to single-cell RNA analytic tools as it will help researchers construct experimental designs with appropriate and accurate power. In Chapter IV, we illustrate how diversity measures can be applied to single-cell RNA-seq data to identify communities of cells (e.g., individuals) or communities of transcripts (e.g., cells) with irregular differences in diversity. Overall, this dissertation demonstrates how single-cell experiments, when appropriately and thoughtfully analyzed, have the potential to fundamentally shift our understanding of cellular biology.
Language: English
Open Access status: green
OpenAlex ID: W7035780116
Cites: 220
Cited by:
Related to: 10