Abstract:
The advances in next-generation sequencing (NGS) technology enabled
rapid and cost-effective whole genome analyses. Nowadays, it is known that
individual organisms have unique genome sequences and that differences
between these sequences are the reason for genetic diversity. Furthermore,
the biomolecular processes of living organisms are steered by genes and the
interplay of their products. Perturbations in these systems often lead to
disease. Thus, one of the major question in biomedical research is how genetic
variations influence gene function, and how these affect underlying biological
pathways and gene interaction networks. One of the most common sources of
genetic diversity are single nucleotide variations (SNVs). So-called Genome
Wide Association Studies (GWAS) as well as expression Quantitative Trait
Locus (eQTL) studies intend to associate SNVs with e.g. disease related
binary or quantitative traits. However, available methods are usually limited
to statistical analyses and previous approaches to improve the interpretation
of the respective results are often insufficient.
The goal of this dissertation was the development of new visual analytical
approaches to assist purely statistical methods in the identification, characterization
and interpretation of SNVs.
Genomic variations, especially SNVs, also play an important role in the
immensely growing field of paleogenetics, where DNA of ancient origin is
compared to modern DNA with the intention to gain insights into evolutionary
history. In this dissertation, a computational pipeline for comparative NGS
analyses of ancient and modern DNA samples has been described. Special
attention was given to the read merging step, which is required to cope
with the quality limitations inherent to ancient DNA (aDNA), in particular
DNA fragmentation and nucleotide misincorporation. In addition, aDNA is
usually only retrievable in low amounts and it is often contaminated with
DNA of modern microorganisms. To solve this issue, a highly economical
microarray-based DNA capturing strategy has been developed for the parallel
detection and enrichment of aDNA from up to 100 different human pathogens.
