Automated quantitative analysis methods for translocation of biomolecules in relation to membrane structures

Biological processes are complex study objects due to their dynamic nature and structural diversity of living organisms. To study dynamic processes statistically, numerous experiments with multiple observations have to be performed, and data have to be analyzed and evaluated. Owing to great technological advances, gigabytes of data are being acquired both in research and industry. Slow and subjective manual analyses are not sufficient anymore, and automated evaluation methods are required. The distribution of biomolecules provides valuable information on a current biological state. The distribution of biomolecules depends on and is influenced by functions of biomolecules, and may thus be used to detect abnormalities. The relatively young research field extit{toponomics} describes the laws of spatial arrangement of molecules. Several evaluation methods have previously been developed, automatized and standardized. However, no standard evaluation methods have been reported to quantitatively analyze such an important biological process like translocation of biomolecules. Translocation processes are vital for living organisms. For instance, substance inclusion into a cell or exclusion from it represent a translocation. Furthermore, signaling biomolecules translocate from the cytoplasm across the nuclear membrane into the nucleus to influence gene and protein expression. Investigating translocation processes may help to understand complex biological functions. It may also be used to analyze signaling events, or may even be employed for diagnostics and therapy monitoring. Manual and case-specific methods for quantitative translocation analysis are known, but fail to be generally applicable. Therefore, I have developed a novel generic automated approach. The method is based on microscopy images of biological samples. I have defined a generic method to quantitatively express distribution of biomolecules in numeric descriptors. Herewith, changes in distribution may be analyzed using different biological samples. Thus, the samples analyzed do not necessarily have to belong to a time series. Furthermore, not only cell cultures, but also tissue samples can be used for the analysis. Evaluations of cell cultures are simpler due to homogeneity and spatial separation of individual objects. However, structural polarity of the cells can be seen only in tissues. I have developed two workflows based on numeric descriptors for the distribution of bio-ewline molecules. The first workflow uses structure detection in images to localize the objects for evaluation. The second workflow avoids this complex operation by a structure-independent information extraction strategy. Both workflows are generic and may be applied to quantify a wide range of translocation processes.