Doctor of Philosophy (PhD)


Biological Sciences

Document Type



The human genome is strewn with repetitive sequence. An early estimate derived from the draft human genome sequence placed this repetitive content at ~45%. More detailed recent analyses have advanced the idea that the human repetitive and repeat derived contribution to the genome may be closer to 66-69%. The most commonly repeated sequence in the human genome is the Alu element. Alus make up 10.6 percent of all human DNA and have expanded to over one million copies in the human genome reproducing through a copy and paste mechanism. New Alu germline insertions are estimated to occur at a rate of 1 in 20 human births. In addition to their insertional impact, Alus have also been associated with various forms of genomic sequence disruptions including inversions, rearrangements, translocations and deletions. Chimeric Alus are frequently located at the breakpoints of these various forms of structural variations. This observation has led to the putative conclusion that chimeric Alus primarily result from the non-allelic homologous recombination between Alu elements. However, little proof is available regarding the actual mechanism(s) that catalyze this activity. This dissertation reveals a newly recognized pattern among human Alu pairs that may provide additional insight into the mechanism(s) driving chimeric Alu formation. After adjusting for directional biases associated with clustering, Alu pairs in the same orientation (direct) outnumber Alu pairs in the opposite orientation (inverted pairs) by over two percent (p<0.05). If this imbalance was generated by deletions resulting from interactions between inverted Alu elements, many chimeric Alus may have formed from the homologous repair of these deletions. This dissertation characterizes the human Alu pair imbalance and constructs an Alu-based model of human genome instability. This model was used to compare the relative instabilities of 50 human deletion-prone cancer genes and 50 randomly chosen genes. Taken as separate groups, the 50 deletion-prone cancer genes were estimated to be 58% more unstable than the 50 randomly chosen genes. This approach to estimating human gene instability may lay the foundation for comparing genetic risks unique to specific individuals, families and people groups.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Batzer, Mark A.