We can define an inbred individual as an individual whose parents are more closely related to each other than two random individuals drawn from some reference population. [. . .]
1.6 Summarizing population structure
We defined inbreeding as having parents that are more closely related to each other than two individuals drawn at random from some reference population. The question that natu- rally arises is: Which reference population should we use? While I might not look inbred in comparison to allele frequencies in the United Kingdom (UK), where I am from, my parents certainly are not two individuals drawn at random from the world-wide population. If we estimated my inbreeding coefficient F using allele frequencies within the UK, it would be close to zero, but would likely be larger if we used world-wide frequencies. This is because there is a somewhat lower level of expected heterozygosity within the UK than in the human population across the world as a whole.
Wright (1943, 1951) developed a set of ‘F-statistics’ (also called ‘fixation indices’) that formalize the idea of inbreeding with respect to different levels of population structure. He defined F XY as the correlation between random gametes, drawn from the same level X, relative to level Y.
[. . .] the reduction in heterozygosity within individuals compared to that expected in the total population can be decomposed to the reduction in heterozygosity of individuals com- pared to the subpopulation, and the reduction in heterozygosity from the total population to that in the subpopulation.
Remedial population genetics for Greg Cochran
From Graham Coop's population genetics notes: