Since man started telling stories he has been captivated by his origins. From Genesis to Gilgamesh, originary tales attempted to explain human development. Modern science – hard and soft – has clearly shared this obsession and has picked up the mantle, replacing wild speculation with a smattering of fact.
That’s where researchers like Eran Elhaik come in. By applying evolutionary genomics to populations, he is able to discern snapshots of people by tracing their geographic movements. When used with human fossils, this biogeograpic approach can decipher aspects of the past by analyzing ancient DNA.
Elhaik developed a novel technology called Geographic Population Structure at his University of Sheffield lab. In a 2014 paper published in Nature titled, “Geographic population structure analysis of worldwide human populations infers their biogeographic origins”, he laid the foundation for the use of the aGPS algorithm in analyzing place of origin through DNA.
From the paper’s abstract:
Here we describe the Geograpic Population Structure algorithm and demonstrate its accuracy with three data sets using 40,000-130,000 SNPs. GPS placed 83% of worldwide individuals within their country of origin.
Now, he has used the same tool to analyze a topic that has tantalized and befuddled the scientific world – ancient Eurasian migrations. It holds the promise of adding many more facts to the discourse.
SCINQ spoke with Dr. Elhaik about aGPS, it’s application, and how he applied it in his most recent study, “From lost empires to modern cities with ancient GPS.”
SCIENTIFIC INQUIRER: Can you give some background on how ancient migrations are tracked normally? What are its shortcomings?
ERAN ELHAIK: When studying past migration event—be it birds, wolves, or humans—the strategy is always the same: obtain a global dataset, find samples that are very similar to the samples of interest, and carry out some analyses to evaluate whether the similarity is due to migration.
Ancient migrations are typically traced using modern paternal or maternal haplogroups. In such cases, haplogroups are identified and dated (e.g., R1a1). Provided this information and their frequencies (ancient or modern) in various locations, a most likely migration path is constructed. This method is problematic since the modern frequencies of haplogroups do not represent the past very accurately.
Ancient DNA opens a window to the past and allows assessing past events more accurately. However, even in this case, it is difficult to infer population movements from haplogroups since it is very difficult to anchor them to certain geographical regions. Ancient autosomal data should be preferred, however it is rarely used to infer migrations. At best, similarity between samples may be used to compare different migration models.
SCINQ: How has genetic technology changed the study of ancient populations? Where does the ancient Geographic Population Structure tool fit into that scheme?
EE: The availability of large amounts of ancient DNA allows us to test complex scenarios that we could not evaluate before.
For example, one of the problems with ancient DNA is that we have uncertainty as to who is local and who is migrant. Here, aGPS does two things: first, it employs ancient autosomal data that varies over time and space and constructs a multi-dimensional genetic-geographic continuum which allows testing if samples can be predicted to their concurrent locations (burial site). This process is refined until a subset of samples has been confirmed to be reliably associated with their burial site and the remaining individuals are the migrants.
Since we know their age, and now we know where they came from, we can develop migration maps. These maps can now be compared with maps constructed using other means, such as parental haplogroup, archeological findings, or linguistic data.
SCINQ: How does the ancient Geographic Population Structure tool work? What are its limitations?
EE: The accuracy of aGPS depends on the comprehensiveness of the aDNA. The more complete the data are over time and space, the more accurate the migration routes would be. At the moment, for example, we cannot see back migrations to North Africa because we don’t have aDNA from this region.
SCINQ: How was your current study designed and what was its objective?
EE: We aim to reconstruct the human migration atlas by developing novel tools that can analyze ancient DNA.
SCINQ: What did it reveal about ancient Eurasian migrations?
EE: This will be reported in details in the paper, but you see some of the questions we studied in here (https://www.sciencedaily.com/releases/2017/05/170528192059.htm).
SCINQ: What are the broader implications for the use of the new tool? Can it replace older methods? Does it have other applications besides how it was used?
EE: aGPS allows testing hypotheses thus far debated only among social scientists using archeological, historical, or linguistic evidence. It is best used in conjunction with data collected from other sources to give the results context.
For example, aGPS will not tell you why people left the Levant after the development of agriculture. Archeological and geological evidence suggests that they depleted the resources of their surrounding and had to move.
SCINQ: How does your current study tie into your past endeavors? How does it influence future lines of inquiry?
EE: aGPS is the latest tool in the serious of GPS tools. The first GPS tool (published in Nature communications in 2014 http://www.nature.com/ncomms/2014/140429/ncomms4513/full/ncomms4513.html and was highlighted in Science http://news.sciencemag.org/archaeology/2014/04/scienceshot-genetic-app-tells-you-where-you%E2%80%99re) predicted the geographical origin of modern day individuals. The second tool, GPS Origins (commercially available at https://homedna.com/gpsorigins), predicted two geographical origins of modern day individuals and inferred their migration routes. Both tools however are limited in their ability to infer ancient migrations, a gap that the current tool aims to bridge. aGPS would allow studying the origin of worldwide populations with a better resolution than available thus far by alternative tools.
SCINQ: What is the one research tool you cannot live without and why?
EE: Our work follows the findings of renowned experts like Luigi Luca Cavalli-Sforza and Sergio Tofanelli who studied the relationships between genetics and geography and length. In that respect our work confirmed and expands their findings. We are also making use of many bioinformatic tools that help us process the data and are proud to contribute our tool to this vibrant community.