Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Interdisciplinary and Graduate Studies

Degree Program

Interdisciplinary Studies with a specialization in Bioinformatics, PhD

Committee Chair

Kalbfleisch, Theodore S.

Committee Co-Chair (if applicable)

Rouchka, Eric C.

Committee Member

Park, Juw Won

Committee Member

Watson, Corey

Committee Member

Gill, Ryan

Author's Keywords

Equine genomics; introgression; ancient introgression; bioinformatics; population genetics


Introgression is one of the main mechanisms that transfer adapted alleles between species. The advantageous variants will get positively selected and retained in the recipient population while rest of the variants undergo negative selection. When analyzing horse genome, two alleles were found in CXCL16 gene, one associated with susceptibility and one with resistance to developing persistent shedding of the Equine Arteritis Virus. The two alleles differ by 4 non-synonymous variants in exon 1 of the gene. Comparison with 3 non-caballine equids (zebras, asses and hemiones) revealed that one haplotype was almost identical to the haplotype found in non-caballines while the other had differences characteristic of 4.5 million years since a common ancestor. Based on this observation, we project that an ancient introgression event occurred between caballine and non-caballine equids. If so, we should be able to find more instances of introgression between these species. We developed a method to identify putatively introgressed segments in the horse genome. It is estimated that non-caballine equids such as zebras and asses diverged from horses between 4 and 4.5 MYA. Genomic analysis of these animals vs. equine reference genome reveals the divergence at both the nucleotide and chromosomal level. Whole genome data for the non-caballine equids when mapped to the caballine (Equus caballus) reference genome show a greater frequency of single nucleotide differences than horses have relative to the same reference. We have created a Likelihood Estimate framework that uses this difference in single nucleotide frequencies to predict whether a haplotype evolved along the caballine or non-caballine lineage. Our results demonstrated that these haplotypes are between 0.5 and 2kb in length and are detectable at a rate of several hundred loci per horse. About 1.1% of the equine genome was introgressed and 64% of the identified putative regions were associated with either structural elements, regulatory regions, or both. These regions were responsible for gene products involved in regulation of response to stimuli, signal transduction, integral components of cell membrane and important metabolism pathways such as purine metabolism and thiamine metabolism. Furthermore, these haplotypes occur at high frequency in the horse population suggesting that they are positively selected by evolution.