Item

Abstract

Genome sequencing projects produce large amounts of data that are stored in
sequence databases. Entries in these databases are annotated using the results
of different experiments and computational methods. These methods usually rely
on homology detection based on sequence similarity searches.
Gene Ontology (GO) provides a standard vocabulary of functional terms, and
allows a coherent annotation of gene products. These annotations can be used as
a basis for new methods that compare gene products on the basis of their
molecular function and biological role.
In this thesis, we present a new approach for integrating the species taxonomy,
protein family classifications and GO annotations. We implemented a database
and a client application,
GOTaxExplorer, that can be used to perform queries with a simplified language
and to process and visualize the results. It allows to compare different
taxonomic groups regarding the protein families or the protein functions
associated with the different genomes. We developed a method for comparing GO
annotations which includes a measure of functional similarity between gene
products. The method was able to find functional relationships even if the
proteins show no significant sequence similarity. We provide results for
different application scenarios, in particular for the identification of new
drug targets.