Date of Award
Doctor of Philosophy (PhD)
Genomics & Computational Biology
In this dissertation, we used single-cell RNA sequencing data from five mammalian tissues to characterize patterns of gene expression across single cells, transcriptome-wide and in a cell-type-specific manner (Part 1). Additionally, we characterized single-cell RNA sequencing methods as a resource for experimental design and data analysis (Part 2).
Part 1: Differentiation of metazoan cells requires execution of different gene expression programs but recent single cell transcriptome profiling has revealed considerable variation within cells of seemingly identical phenotype. This brings into question the relationship between transcriptome states and cell phenotypes. We used high quality single cell RNA sequencing for 107 single cells from five mammalian tissues, along with 30 control samples, to characterize transcriptome heterogeneity across single cells. We developed methods to filter genes for reliable quantification and to calibrate biological variation. We found evidence that ubiquitous expression across cells may be indicative of critical gene function and that, for a subset of genes, biological variability within each cell type may be regulated in order to perform dynamic functions. We also found evidence that single-cell variability of mouse pyramidal neurons was correlated with that in rats consistent with the hypothesis that levels of variation may be conserved.
Part 2: Many researchers are interested in single-cell RNA sequencing for use in identification and classification of cell types, finding rare cells, and studying single-cell expression variation; however, experimental and analytic methods for single-cell RNA sequencing are young and there is little guidance available for planning experiments and interpreting results. We characterized single-cell RNA sequencing measurements in terms of sensitivity, precision and accuracy through analysis of data generated in a collaborative control project, where known reference RNA was diluted to single-cell levels and amplified using one of three single-cell RNA sequencing protocols. All methods perform comparably overall, but individual methods demonstrate unique strengths and biases. Measurement reliability increased with expression level for all methods and we conservatively estimated measurements to be quantitative at an expression level of ~5-10 molecules.
Dueck, Hannah R., "Single-Cell Gene Expression Variation as A Cell-Type Specific Trait: A Study of Mammalian Gene Expression Using Single-Cell RNA Sequencing" (2015). Publicly Accessible Penn Dissertations. 1692.