This thesis covers novel applications of concepts from communications engineering to problems in molecular biology. In the first part the focus is placed on applying source coding techniques to reduce the storage requirement of multiple genome alignment datasets used in comparative genomics. A highly efficient lossless compression algorithm using well established models of genome evolution and binary image compression techniques is introduced. The second part studies parallels between sequence specific protein binding on the molecular level and threshold based marker synchronization over noisy channels. The engineering concept of threshold based marker synchronization over noisy channels is revised and extended. Binding site models and in silico inference techniques are reviewed using information theory. Synchronization properties of molecular markers are analysed and evidence for selection pressure towards good markers is found.
«