Aug 18, 2021

Save for later NEW YORK – A new study has shed light on the genetic relationship between Finns and Estonians, Northern European populations that, while separated by the Gulf of Finland, have languages that are closely related, suggesting recent common ancestry. The effort, highlighted in an American Journal of Human Genetics paper published online Wednesday , relied on Estonian Biobank data, the sequencing of archaic genomes, and diverse analytical methods to assess the relatedness of the kindred populations. Researchers at the Institute of Genomics at the University of Tartu in Estonia, the department of human genetics at KU Leuven in Belgium, University College London, Cambridge University, the British Library, Cornell University, the University of Padova in Italy, and Italy's National Research Council contributed to the effort. According to lead author Toomas Kivisild, a population geneticist at KU Leuven, the study is part of "broader and longer-term research into the genetic origins of Finno-Ugric populations" undertaken by the Institute of Genomics in Tartu. "Using ancient DNA we can directly address questions about when specific ancestry components appear in time," Kivisild said in an email. Tracing the split between the Finnish population and Estonian population is of interest to the researchers, as previous studies have shown their common ancestors reached the Baltic region in the first millennium BC. Yet the closeness of the two languages, in addition to other historical events, such as evidence of Finnish settlements in northern Estonia in the 17th and 18th centuries, suggest that the two populations diverged more recently. The researchers have therefore sought to pinpoint when this split occurred. One issue confronting the project was determining the efficacy of imputing low-coverage ancient genomes. As part of the effort, they used the Illumina NextSeq 500 platform to sequence the remains of two individuals from medieval burials in south and northeast Estonia and used a sample from Cambridge as a control. The results were compared to genotyping array data on 143,000 modern-day Estonians from the Estonian Biobank, as well as 99 Finns and 14 imputed ancient genomes from Estonia. By searching for long identity-by-descent segments shared among the populations, the authors found high levels of individual connectedness between Finns and Estonians for the last eight centuries. The results led the researchers to posit there was a migration event from north Estonia to Finland sometime between the years 800 AD and 1000 AD, before the expansion of the Finnish founder population. Finland currently has a population of 5.5 million but is believed to have had an original population of about 50,000 people. Kivisild said that the researchers used long-allele interval sharing tools similar to those used by genetic ancestry companies such as 23andMe or Ancestry to detect relatives and found high numbers of genetic connections between Finns and Estonians that date back at least 800 years, meaning they are not inherited from recent genealogical relationships. "In contrast to the patterns of allele frequencies where Estonians are most similar to [Latvians and Lithuanians], their southern neighbors from the Baltic countries, the sharing of the long-allele intervals appears to be the highest with Finns," said Kivisild. He noted that since the level of these IBD segments peaks in the 800-year-old ancient genome samples, rather than pre-Roman, Iron, and Bronze Age samples from Estonia, the researchers believe such a migration occurred. "That, according to linguists, explains why Estonian and Finnish languages are so similar to each other as they are," Kivisild said. Kivisild said that the study also suggests that long shared segments can provide a computationally tractable way to detect fine-scale structure in large cohorts. The work will continue. Kivisild said that his team is now exploiting genetic relationships in medieval sites across Estonia and are awaiting similar data from Finland as well as other neighboring countries. This will help the researchers contextualize the data and better understand the past, he said.