Using Visual Phenotypes to Dissect Sequence-Function Relationships and Complex Drug Responses

Abstract

Cellular morphology is a potent indicator of cellular function and dysfunction, but the relationships between morphology, genetic variants, and cellular state remain incompletely understood. In this thesis, I describe a method called Visual Cell Sorting, which can be used to systematically characterize cellular morphologies and other visual phenotypes of interest. In a Visual Cell Sorting experiment, automated imaging and phenotypic analysis directs selective illumination of Dendra2, a photoconvertible fluorescent protein expressed in live cells; these photoactivated cells are then isolated using fluorescence-activated cell sorting. Visual Cell Sorting can be used to characterize hundreds of genetic variants according to a visual phenotype and to discover genes that are responsible for maintaining homeostasis in response to drug treatment. Visual Cell Sorting’s greatest strength is that a variety of downstream assays can be performed on the separated cells, which together can characterize a morphologic phenotype in a multimodal and systematic fashion.