Admixture in New Zealand isolates
In founder events involving a few initial founders, the small effective
population size will result in strong effect of genetic drift and
increased linkage disequilibrium, often resulting in large changes in
allele and haplotype frequencies. Thus, founder events complicate
estimates of the relationship between populations using allele
frequency-based methods, such as Admixture. However, the
haplotype structure contains information that may be used to identify
the relationship between populations. To utilize the haplotype structure
to infer the relationship between the New Zealand population and the
European and Japanese populations we performed a local ancestry
assignment using PCAdmix (Brisbin et al. 2012) with both
Europe and Japan as ancestral groups and the New Zealand isolates as an
admixed population. PCAdmix uses phased haplotypes derived from
both reference panels and the admixed isolates to infer the ancestry of
genomic windows. We used Beagle 4.0 (Browning & Browning 2007)
to phase the data for each population, without imputation, using default
settings. The local ancestry was inferred with PCAdmix, which
uses principal components analysis in windows of 10 SNPs along the
genome to identify ancestry from a set of predefined potential ancestral
populations. Thus, in the case of the two individuals from New Zealand,
the populations from Europe and Japan were defined as the ancestral
populations.