Ancient DNA (aDNA) has shown dramatic technological and methodological advancements since the rst short, ancient mitochondrial DNA sequence was retrieved in 1984. Today, in particular thanks to high-throughput DNA sequencing (HTS) technologies, complete genomes from a wide variety of species, including unknown and/or extinct lineages, have been characterized. In addition, HTS data can also unveil pre-mortem DNA methylation proles by tracing non-random DNA damage patterns. However, a statistical model that accounts for sequencing uncertainties and provides a cross-sample comparable methylation estimate is missing. In this PhD, we developed DamMet, the rst software to recover maximum likelihood estimates of cytosine methylation from ancient specimen. The statistical model incorporates sequencing uncertainties, marginalize over true variants in dinucleotide space, and accounts for dierential deamination rates at methylated and unmethylated sites. We show that the regional methylation estimates are directly comparable to those retrieved from fresh tissue. This opens for the opportunity to trace the methylome through time, including major changes such as the neolithic transition and animal domestication processes. The genetic makeup of domesticated horses was the other main focus of this PhD. The horse had a huge impact on the human history by, among others, changing our mobility and agricultural productivity as well as the way we made war. However, the trajectory of horse domestication from the rst domesticates until present-day breeds has proven dicult to disclose based solely on patterns of modern genetic variation and archaeological evidence. Leveraging on the largest non-human HTS aDNA data set, this PhD aimed at disentangling the genetic history of horse domestication. We dismissed two proposed domestication centers for the ancestors of present-day horse breeds. One candidate, the domesticated Botai horses from Northern Kazakhstan, was instead the direct ancestors to the Przewalski's horses, previously thought to be the last extant wild lineage. The second candidate on the Iberian Peninsula turned out to be a novel horse lineage. We also showed that ancestors to present-day horses were reshaped in the 7th-9th century AD, following the Islamic conquest and that most present-day breeds show signicant evolutionary rootsin Sassanid Persia.