Date of Completion


Embargo Period



purifying selection, selection for function, dN/dS, selfish genes, robustness

Major Advisor

Dr. Johann Peter Gogarten

Associate Advisor

Dr. Paul Lewis

Associate Advisor

Dr. Victoria Robinson

Associate Advisor

Dr. Daniel Gage

Associate Advisor

Dr. Joerg Graf

Field of Study

Genetics and Genomics


Doctor of Philosophy

Open Access

Open Access


The research presented in this thesis attempts to address research questions related to the role of natural selection in the evolution of bacterial genes not expressed for function and in building mutational tolerance to translational errors. Studies on evolution of protein coding DNA sequences have provided the evidence for a current paradigm in evolutionary biology: only functional genes are undergoing selection against the deleterious effects of allele variants (purifying selection). I provide evidence that similar footprints of selection can be detected in genes that are not normally expressed for function during the bacterial life cycle. Using simulations for DNA sequence evolution, I demonstrate statistically significant deviations from neutral evolution for the studied genes. I suggest that purifying selection affects both functional and non-functional genes. I propose this might be caused by the dominant toxic effects of low level translation of mutated products in bacteria, due to misfolding and misinteraction. Natural selection also acts to remove the effects of translational errors. Stop codon readthrough events are more likely to have major structural and functional effects than simple nucleotide changes. Recent research has shown that strength of selection experienced by protein-coding genes is positively correlated with the level of gene expression. Expression of 3’ untranslated regions (3’ UTRs) carries with it the influence of natural selection on elongated products. I show that, for the subset of highly expressed genes analyzed, 3’ UTRs in Escherichia coli genomes display features normally associated with coding regions, indicating tolerance to effects of translational errors.

PhD_Dissertation_SOmer_Associated_File_1.xlsx (246 kB)
Gene identification numbers