UCLA

The genetics of gene regulation is very challenging to study at high resolution. Mapping mRNA levels as complex traits is a powerful approach. High-resolution mapping of gene expression traits (eQTL) is possible in yeast. To the extent that this has been studied, genetic effects on expression appear to depend greatly on the environmental and physiological state context in which they are measured. Major open questions relate to the way eQTL maps change as a function of physiological state and environment and how these changes are related to the realization of organismal and fitness traits. Single-cell RNAseq-based eQTL mapping allows high-resolution studies of genetic gene regulation across a litany of environmental conditions, facilitates physiological state decomposition prior to mapping, and enables experimental designs that increase interpretability via perturbation.In this dissertation we optimize a single-cell RNAseq approach to yeast eQTL mapping so that it is suitable for a wide range of conditions and cellular properties. Next, we use single-cell RNAseq-based eQTL mapping to conduct high-resolution eQTL mapping across two perturbation time courses. With single-cell RNAseq, we were able to conduct acute perturbation eQTL time points as well as to resolve physiological state heterogeneity. We conduct genetic mapping of occupancy in these states and identify many genetic loci. We conduct eQTL mapping within each physiological state subset as well as in a way that pools information from all cells in each time point to maximize statistical power. We observe that these maps change dramatically as a function of environment and physiological state. Finally, we focus on eQTL hotspots and the biological processes that they likely influence. We observe that expression hotspots not only change as a function of physiological state but that the linking transcripts are highly related to the processes occurring in the contexts where they are measured. We find many overlaps between expression QTL hotspots and fitness trait QTL identified in the same populations. We use these overlapping signals to construct hypotheses regarding causal relationships between expression hotspots, organismal traits, and fitness traits.