Date of Award
Spring 2023
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Genetics
First Advisor
Hammarlund, Marc
Abstract
Nervous systems are complex networks, which requires a diverse set of neurons to detect, process, learn from, and respond to external stimuli. This neuronal diversity is directly tied to diversity of gene expression networks across the nervous system. Here I present my work, as part of the CeNGEN consortium, mapping gene expression across the entire nervous system in the hermaphroditic nematode C. elegans. I outline a strategy for sequencing RNA from individual neurons using bulk RNA-seq after cell sorting, and demonstrate a species-specific reagent for ribosomal depletion for ultra-low input RNA-seq library preparation. I describe a collaborative effort to sequence cells from all 118 neuron classes in the L4 C. elegans using single cell RNA-seq (scRNA-seq), and key findings for gene regulation, sensory modalities, and genetic drivers of functional connectivity. I describe a complementary bulk RNA-seq dataset for 41 neuron classes, and detail a computational method for integrating bulk RNA-seq data with matched scRNA-seq replicates that reduces false positives in gene detection and differential gene expression. I also characterize a noncoding RNA map for those 41 neurons in the bulk RNA-seq dataset, many of which are missed using scRNA-seq techniques. Further leveraging the combined power of scRNA-seq and bulk approaches, I demonstrate a flexible iterative subtraction method for removing contaminating counts from bulk RNA-seq datasets using an scRNA-seq reference, and show that it improves gene detection beyond current cell-type specific expression tools. Building on previous studies of scRNA-seq detection rates, I describe a simple model for denoising aggregated pseudobulk replicates using the proportion of cells that detect a gene, and show that it provides a robust measure for differential gene expression analysis, with fewer false positives. Finally, I show how integrating bulk RNA-seq after subtraction with denoised scRNA-seq pseudobulk replicates marries the strengths of both measurements, and improves accuracy in gene detection and differential expression across neuron types in C. elegans. These findings provide important new tools and strategies for linking gene expression to neuronal shape and function.
Recommended Citation
Barrett, Alec, "Integrative Analysis of Bulk and Single Cell RNA-seq modalities in the C. elegans Nervous System" (2023). Yale Graduate School of Arts and Sciences Dissertations. 1081.
https://elischolar.library.yale.edu/gsas_dissertations/1081
