UBC Theses and Dissertations
Computational approaches to the study of genomic roles of repeated DNA sequences Van de Lagemaat, Louie Nathan
Repeated sequences make up nearly half of the bulk of mammalian genomes and vary widely in structure and function. This thesis describes computational approaches for assessment of interaction of repeats and their host genomes. Following public release of the human genome sequence, initial investigations focused on overall distributions of retroelements with respect to sequence composition and genic position. Exclusion of various retroelements from regions both within and surrounding protein-coding genes suggested selection against the presence of these elements. Directional biases of accepted retroelements in these regions further supported this notion. Directional biases are understood to reflect differential mutagenicity by sequences in one direction vs. the other. To examine the relationship between protein-coding genes and mobile elements further, mappings of genomic transposable elements in the human and mouse genomes were examined in relationship to positions of all exons of protein-coding gene mRNAs. I found that approximately one quarter of mRNAs of protein-coding genes harbor sequence contributed by transposable elements. The fact that transposable element sequence is most often found in untranslated regions (UTRs) suggests a highly significant role for these sequences in modulation of translation efficiency in addition to roles in transcription. Further investigations used directional biases of retroelements in transcribed regions in humans and mice to show that transposable elements transcribed by RNA polymerase II (pol II) exert varying effects upon insertion, depending on the sequence of the element. Finally, a bioinformatic study done on global insertion patterns of retroelements since human-chimpanzee divergence revealed that some transposable elements polymorphic for presence or absence in primate genomes actually represented deletions rather than de novo insertions. These deletions were flanked by short tracts of identical sequence, suggesting deletion by recombinational mechanisms. The relative rarity of these events lends support to the assumed stability of transposable element insertions while illustrating the recombinational activity of even low-copy, nonadjacent, short repeated sequences, such as those found flanking transposable element insertions. In summary, these bioinformatic studies lend insight into the biological roles and genomic effects of mammalian genomic repeats, especially transposable elements.
Item Citations and Data