Inference of ancestry: constructing hierarchical reference populations and assigning unknown individuals

Table 2 Re-assignment characteristics of unknown samples to reference clusters defined in the hierarchical decomposition of a dataset by structure

		Raw re-assignment						Post-clustering adjustment re-assignment
Level of resolution	No. clusters	Median cluster size	RASR ^a	No. clusters > 90% RASR	Criterion	R²	No. clusters	Median cluster size	RASR	No. clusters > 90% RASR
1	1	1,056	1.000	-	-	-	-	-	-	-
2	7	86	0.995	6	IMM ^b	0.900	7	86	0.999	6
					Cluster size	0.427
3	30	19	0.939	21	IMM	0.931	29	18	0.960	23
					Cluster size	0.216
4	78	4	0.818	40	IMM	0.943	72	3	0.823	39
					Cluster size	0.296
5	113	3	0.633	47	IMM	0.898	87	3	0.700	42
					Cluster size	0.662

Re-assignment success rates were high over the first several levels of resolution in the analysis. The decrease in re-assignment success over the course of the hierarchy was primarily attributed to an increase in the number of samples exhibiting mixed membership properties among multiple clusters. Despite the decline in overall success, a large proportion of the defined clusters at each tier demonstrated highly successful re-assignment rates. These high-performance clusters have the potential to provide highly informative assignments for truly unknown individuals in ancestry testing procedures. Additionally, post-clustering adjustments, wherein certain outlier samples were systematically relocated, had the effect of increasing success rates at all levels of resolution.
^a RASR = Re-assignment success rate
^b IMM = Individual mixed membership (in multiple clusters)

ISSN: 1479-7364