Answering the big WHEN and WHERE questions in paleogenomics - with applications to the histories of Vikings and horses

Reconstructing past events related to the genetic diversity of a species and their struggle with diseases is the Holy Grail of population and medical genetics because only by understanding the past can we understand why modern populations are genetically different and how to protect them from diseases.

In the past decade, sequencing ancient genomes and using them to improve reconstructions became feasible, resulting in the accumulation of thousands of human and animal genomes.

These genomes shed light on past populations and our ignorance about them, partially because the rate of data generation surpasses the rate of method development. To harness these genomes for population and medical studies we must first know WHEN and WHERE are they from.

Unfortunately, less than half of all the genomes are radiocarbon dated and no biogeographical tools exist, which questions the usefulness of paleogenomics. Therefore, developing reliable bio-dating and bio-localizing tools is a top priority in the field.

Motivated by our past and preliminary work, we will develop machine-learning-based tools to genomically date samples and “rescue” thousands of undated genomes. We will also predict their geographical origins.

We will then apply these and other tools to reconstruct the history of the Vikings and trace the genetic origin of modern horse domestication, which will not only demonstrate the usefulness of the tools but address two of the most challenging questions in paleogenomics.