Year: 2007
Type: dissertation
Abstract: In this thesis exploratory data analysis methods have been developed for analyzing genomic data, in particular human endogenous retrovirus (HERV) sequences and gene expression data. HERVs are remains of ancient retrovirus infections and now reside within the human genome. Little is known about their functions. However, HERVs have been implicated in some diseases. This thesis provides methods for analyzing the properties and expression patterns of HERVs.\n\nNowadays the genomic data sets are so large that sophisticated data analysis methods are needed in order to uncover interesting structures in the data. The purpose of exploratory methods is to help in generating hypotheses about the properties of the data. For example, by grouping together genes behaving similarly, and hence presumably having similar function, a new function can be suggested for previously uncharacterized genes. The hypotheses generated by exploratory data analysis can be verified later in more detailed studies. In contrast, a detailed analysis of all the genes of an organism would be too time consuming and expensive.\n\nIn this thesis self-organizing map (SOM) based exploratory data analysis approaches for visualization and grouping of gene expression profiles and HERV sequences are presented. The SOM-based analysis is complemented with estimates on reliability of the SOM visualization display. New measures are developed for estimating the relative reliability of different parts of the visualization. Furthermore, methods for assessing the reliability of groups of samples manually extracted from a visualization display are introduced.\n\nFinally, a new computational method is developed for a specific problem in HERV biology. Activities of individual HERV sequences are estimated from a database of expressed sequence tags using a hidden Markov mixture model. The model is used to analyze the activity patterns of HERVs.
Source: Aaltodoc (Aalto University)
Author: Merja Oja
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W2580572024