Fig. 8

The graphical explanation of the method for constructing the datasets. For example, two resource populations, A and B, were sampled to investigate genetic diversity. Resource population A was used to upgrade population B; i.e. there is asymmetric gene flow from A to B. To create a diversity sample consisting of random animals with relationships not stronger than the average in the resource population, researchers use written and verbal information. After sampling and genome-wide genotyping, some closely related samples could and should be excluded from further analyses requiring unrelated individuals (population structure and phylogenetic analyses). Successive exclusion of one from the most closely related pair of animals is commonly used approach and resulted in a one-step improved diversity sample (Alp1Step), as illustrated in the first path of population B. Foreign or admixed individuals are prone to weaker relationships within the sample, and consequently, after the one-step procedure, they remain in the improved diversity sample. A preceding outlier test (path 2 for B) will decrease the inflated diversity within and increase the diversity between two-step improved diversity samples (Alp2Step, Euro2Step and Glob2Step)