Anmerkungen:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Beschreibung:
The past decade in molecular biology was characterised by a multitude of genome-wide studies which revealed fascinating insights into the complexity of genomic organization in all kingdoms of life. Surprisingly, a large extend of the transcriptional output consists of non-coding RNAs (ncRNAs), transcripts not being translated into proteins. The sheer amount of functional identified ncRNAs is just overwhelming and high-throughput sequencing technologies produce genomic and transcriptomic sequence data with an ever increasing rate. However, a precise functional annotation of the majority of RNA transcripts remains a challenge. Comparative bioinformatic approaches are commonly used for large-scale functional analysis and annotation of this immense amount of sequence data. The close linkage of sequential and structural properties in ncRNAs necessitates comparison approaches with usually high computational complexity, which in turn makes even small-sized studies often infeasible for existing approaches. In my thesis, I address these challenges and describe efficient and accurate computational methods for the comparative analysis of RNAs based on their sequence and structure. In the first part of this thesis, I present ExpaRNA, a pairwise RNA comparison approach that uses a novel similarity measure based on exact matching substructures (EPMs). In contrast to previous methods, ExpaRNA detects conserved substructures during the RNA comparison, which is a key feature to detect functional motifs. Instead of generating a full sequence-structure alignment, the developed dynamic programming algorithm efficiently computes an optimal, non-crossing arrangement of matching substructures. I have shown that this longest common subsequence of exact pattern matchings is in good agreement with existing comparison approaches, but can be obtained in a fraction of runtime. In addition, I investigated a generally applicable approach to speedup ressource-expensive sequence-structure alignment methods by using ExpaRNA’s EPM set as anchor ...