Changes in the fine-scale genetic structure of Finland through the 20th century

Abstract

Information about individual-level genetic ancestry is central to population genetics, forensics and genomic medicine. So far, studies have typically considered genetic ancestry on a broad continental level, and there is much less understanding of how more detailed genetic ancestry profiles can be generated and how accurate and reliable they are. Here, we assess these questions by developing a framework for individual-level ancestry estimation within a single European country, Finland, and we apply the framework to track changes in the fine-scale genetic structure throughout the 20th century. We estimate the genetic ancestry for 18,463 individuals from the National FINRISK Study with respect to up to 10 genetically and geographically motivated Finnish reference groups and illustrate the annual changes in the fine-scale genetic structure over the decades from 1920s to 1980s for 12 geographic regions of Finland. We detected major changes after a sudden, internal migration related to World War II from the region of ceded Karelia to the other parts of the country as well as the effect of urbanization starting from the 1950s. We also show that while the level of genetic heterogeneity in general increases towards the present day, its rate of change has considerable differences between the regions. To our knowledge, this is the first study that estimates annual changes in the fine-scale ancestry profiles within a relatively homogeneous European country and demonstrates how such information captures a detailed spatial and temporal history of a population. We provide an interactive website for the general public to examine our results.

FSTruct: AnFST‐based tool for measuring ancestry variation in inference of population structure

“…This statistic is maximal if each individual possesses a permutation of the membership vector (1,0,…,0) and minimal if all individuals possess membership vector . Kerminen et al (2021) evaluated the Shannon entropy applied to individual‐level membership vectors, assessing variation in time in the Shannon entropy for study participants with different birth years. Whereas both the clusteredness statistic of Rosenberg et al (2005) and the Shannon entropy statistic of Kerminen et al (2021) consider variability of the ancestry coefficients of single individuals, our measure examines variability of ancestry coefficient vectors across individuals.…”

FSTruct: anFST-based tool for measuring ancestry variation in inference of population structure

“…This statistic is maximal if each individual possesses a permutation of the membership vector (1, 0, …, 0) and minimal if all individuals possess membership vector . K erminen et al . (2021) evaluated the Shannon entropy applied to individual-level membership vectors, assessing variation in time in the Shannon entropy for study participants with different birth years.…”

“…Whereas both the clusteredness statistic of R osenberg et al . (2005) and the Shannon entropy statistic of K erminen et al . (2021) consider variability of the ancestry coefficients of single individuals, our measure examines variability of ancestry coefficient vectors across individuals.…”

“…Kerminen et al (2021) evaluated the Shannon entropy applied to individual-level membership vectors, assessing variation in time in the Shannon entropy for study participants with different birth years. Whereas both the clusteredness statistic ofRosenberg et al (2005) and the Shannon entropy statistic ofKerminen et al (2021) consider variability of the ancestry coefficients of single individuals, our F ST /F max ST measure examines variability of ancestry coefficient vectors across individuals. Thus, for example, comparing individuals in corresponding matrices in Figure4Aand 4D, clusteredness increases (and Shannon entropy decreases) as the membership of the highest-membership cluster increases from Figure4Ato Figure4D.…”

Biallelic expansion in RFC1 as a rare cause of Parkinson’s disease

“…Samples were received from 161 subjects and three patients were found with biallelic RFC1 (AAGGG) exp giving a frequency of 1.9 % (0.4–5.4 %; 95 % confidence interval). Intriguingly, the prevalence of PD is higher in North Karelia than elsewhere in Finland 10 , which may at least partly be attributed to factors related to geographically clustered genetic structure of the Finnish population 11 . Biallelic (AAGGG) exp in RFC1 may thus be one of the most common genetic causes of PD in Finland, at least in North Karelia.…”