|
|
Segmentation of Multi-Dimensional Multi-Parametric Microscopic Data of Biological Samples Using Convolutional Neural Networks
Backová, Lenka ; Benda, Aleš (advisor) ; Schätz, Martin (referee)
Multi-parametric highly dimensional images have become a standard way of imaging biological samples. To quantify results from these images, segmentation must be often applied first. However, due to the underlying shortcomings of the fluorescence microscopy of biological samples, i.e. low signal-to-noise ratio, convolutional neural networks have become widely used for automatization of the segmentation. Convolution neural networks showed to be versatile in their potential uses and able to segment complex images. In this work, we utilise neural network U-Net for segmentation of images, which contain not only intensity information, but fluorescence excited state lifetime information as well. We try different representations of the data to assess, whether the added information of the pixel values leads to improved performance. We present an application of the segmentation results with phasor analysis to study the fertility of mice sperm. 1
|
|
|
Bioinformatical analysis of the complex multidimensional microscopy datasets
Backová, Lenka ; Černý, Jan (advisor) ; Čapek, Martin (referee)
Microscopy is embedded in the history of life sciences and vice versa. Recent advances in the field present new challenges as new revolutionary technologies arise. Sample prepa- ration, microscope operation and data analysis have become particularly demanding re- quiring specific interdisciplinary expertise. Bioimaging data analysis is computationally demanding, as microscopy technologies can easily acquire data of exceptional size, often in terabytes. Correct analysis requires computer vision knowledge, as well as knowledge of studied biological systems and last, but not least deep understanding of microscopy technology. Tools available for the analysis of the imaging data vary from open-source customizable software with a coverage of multiple tasks to a task specific proprietary software. To choose the best tools for the analysis, analysts should know their options and tasks at hand. In bioimage analysis the tasks needed to be employed depend on the desired outcome and the acquisition technology. Amongst the possible tasks to con- sider belong deconvolution, segmentation and registration. Amount of approaches and algorithms available is progressively growing, resulting in a complex field, difficult to be easily familiar with. My thesis covers different microscopy technologies with emphasis on...
|
|
|
Study of the differences in the architecture of the binding pockets of two major MDR pumps of yeast Saccharomyces cerevisiae, Pdr5p and Snq2p, using their common substrates
Backová, Lenka ; Gášková, Dana (advisor) ; Krůšek, Jan (referee)
Multidrug resistance (MDR) is responsible for the decrease in drug effectiveness on pathogenic microorganisms or tumours. One of the mechanisms of multidrug resistance is drug transport out of the cell (efflux) by membrane transporters - pumps. Main MDR pumps of a yeast species Saccharomyces cerevisiae are Pdr5p and Snq2p, who share high amino acid sequence identity. This thesis focuses on the differences of these pumps, their binding pockets and their arrangement. The binding pocket of Pdr5p is better researched and comparing the results with those of pump Snq2p leads to broader knowledge about the binding pocket of Snq2p. We use disc diffusion assay to determine common substrates of both pumps, ketoconazole and bifonazole. These substrates are used in potentiometric fluorescent probe diS-C3(3) assay. Results of these experiments lead us to the findings that the binding pocket of Snq2p has multiple binding sites. Binding pockets of pump Pdr5p and Snq2p differ in binding sites and their conformation. However, the conformation of both pumps is dynamic, which has been shown after the addition of glucose to supply the pumps with energy. 1
|
guest ::
