Abstract
Following the increasing availability of DNA‐sequenced data, the genetic structure of populations can now be inferred and studied in unprecedented detail. Across social science, this innovation is shaping new bio‐social research agendas, attracting substantial investment in the collection of genetic, biological and social data for large population samples. Yet genetic samples are special because the precise populations that they represent are uncertain and ill‐defined. Unlike most social surveys, a genetic sample's representativeness of the population cannot be established by conventional procedures of statistical inference, and the implications for population‐wide generalisations about bio‐social phenomena are little understood. In this paper, we seek to address these problems by linking surname data to a censored and geographically uneven sample of DNA scans, collected for the People of the British Isles study. Based on a combination of global and local spatial correspondence measures, we identify eight regions in Great Britain that are most likely to represent the geography of genetic structure of Great Britain's long‐settled population. We discuss the implications of this regionalisation for bio‐social investigations. We conclude that, as the often highly selective collection of DNA and biomarkers becomes a more common practice, geography is crucial to understanding variation in genetic information within diverse populations.
Highlights
Recent social scientific interest in bio-social relations – the ways in which individual biology interacts with social environments to produce specific patterning of health and social outcomes – have spurred more efforts to take into account biological information about individuals, including their DNA
We develop a three-stage approach to infer the geography of fine population structure and draw conclusions with regard to the strategic role of geography in the wider inquiry into bio-social relations
A triangulated geography of population structure in Great Britain We use Regional Integrity values at various levels of granularity to derive a synthesised picture of Great Britain’s geography of population structure, assembling the isonymy groups and genetic clusters that result in the highest Regional Integrity
Summary
Recent social scientific interest in bio-social relations – the ways in which individual biology interacts with social environments to produce specific patterning of health and social outcomes – have spurred more efforts to take into account biological information about individuals, including their DNA. The similarity of surname profiles between areas, which is termed ‘isonymy’, can be used to classify areas into any number of k groups Combining these approaches, we can link the genetic cluster of a POBI sampled individual with the isonymy group based on the residence of the participant’s grandparents. We calculate local concentrations of surnames in 1881 parishes and compare each pair of spatial units with respect to similarity of their surname compositions This logic follows Lasker’s (1977) application of isonymy – ‘the recurrence of the same surnames in different ancestral lines in the same pedigree’ (Lasker 1969, 309) – to estimate relatedness between or genetic similarity of populations based on the relative frequency of common surnames: gij 1⁄4. A similarity index, the Adjusted Rand (Hubert and Arabie 1985), is used to measure the correspondence between two alternative partitions of the POBI sample: one partitioned by genetic cluster, and the second by isonymy group of the local area.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
