Mother-bud Detection and Classification in Yeast Cells
Name
Danver Hans Värv
Abstract
In different organisms, asymmetric cell division is a conserved process in which asymmetric inheritance of cellular components gives rise to two cells with different characteristics. In budding yeast, Saccharomyces cerevisiae, after the asymmetric cell division two cells are generated: mother and daughter. Fluorescence microscopy has been widely used to study such microorganisms and hence study the process of cell division. Advances in microscopy have increased output data volumes, making the manual processing of such images expensive. Therefore, developing a pipeline to analyze microscopy images coming from screening experiments would have a high practical impact. In this work, we built a deep-learning-based pipeline for detecting budding (dividing) yeast cells and distinguishing bud from mother cells during yeast division. The final goal is to study if older proteins are being retained on the mother side and whether newly synthesized proteins are inherited towards the bud. The results show that the pipeline was able to detect the cells that are about to divide with an accuracy of 70.42%. Furthermore, 87.72% of the mothers and buds were accurately classified.
Graduation Thesis language
English
Graduation Thesis type
Bachelor - Computer Science
Supervisor(s)
Mohammed A. S. Ali, Karla Juárez Núñez
Defence year
2022