A Global Analysis of Molecular Fluctuations Associated with Cell Cycle Progression in Saccharomyces Cerevisiae PDF Download

Author: Benjamin Thomas Grys
Publisher:
ISBN:
Category :
Languages : en
Pages : 0

Get Book Here

Book Description
Gene and protein expression, turnover, and localization are imperative for cell cycle progression. However, there has been no systematic study of multi-level regulatory events throughout the cell cycle in eukaryotes. To address this void, I developed a pipeline for quantifying changes in protein concentration and localization over the course of the budding yeast cell cycle. This pipeline combines Synthetic Genetic Array technology, high-throughput fluorescence microscopy of a collection of strains expressing Open Reading Frame-Green Fluorescent Protein fusions, and sophisticated deep learning techniques to generate and analyze cell cycle-specific image data for ~75% of the proteome. In developing this pipeline, I demonstrated that the application of deep learning to biological image data can overcome pitfalls associated with conventional machine learning classifiers, including improved performance at classifying subcellular protein localization as well as transferability to diverse image-sets with minimal tuning and training. I used this optimized pipeline to acquire and analyze >123,000 images of ~20 million live cells. I used a neural network (CycleNET) to classify single cell images by cell cycle position, and a second neural network (DeepLoc) to quantify the localization of proteins in 22 unique localization classes. I optimized statistical scoring metrics to identify 825 proteins with fluctuating levels during cell cycle progression, and 405 proteins that change in localization. Different cell cycle stages featured significant movement of proteins between subcellular compartments, including cell cycle-specific turnover of ribosomal subunits and their regulators at the vacuole in early mitosis, a novel observation that may reflect a new mechanism for ensuring the presence of high quality translational machinery during cell cycle progression. I combined these proteomics datasets with new cell cycle-specific gene expression and translational efficiency data, generated by RNA sequencing and ribosome profiling, respectively. Integrating these datasets allowed me to identify new control mechanisms for known cell cycle regulators, implicate new genes in the control of cell cycle progression, and reveal broad trends about how cells leverage different levels of regulation for different groups of genes. Finally, I demonstrated that the integration of my four cell cycle-specific datasets affords power in predicting cell cycle-related functions of uncharacterized and unannotated genes.