Characterization of a Bayesian genetic clustering algorithm based on a Dirichlet process prior and comparison among Bayesian clustering methods

Table 5 Effect of λ on the behavior of DPART (number of alleles was 2)

Ancestral allele freq. and number of loci	{0.5, 0.5} × 50 loci	{0.8, 0.2} × 200 loci	{0.5, 0.5} × 50 loci + {0.8, 0.2} × 50 loci
Mean major allele frequency	0.621 ± 0.088	0.802 ± 0.114	0.711 ± 0.137
λ = 6	0.067 (83)	0.500 (0)	0.440 (12)
	2.09	1.00	1.12
λ = 3	0.085 (73)	0.470 (6)	0.176 (66)
	2.31	1.06	1.66
λ =	0.325 (36)	0.050 (90)	0.208 (59)
	1.44	1.90	1.59
Inferred (unique)	0.104 (70)	0.080 (84)	0.063 (89)
	2.22	1.84	1.89
Inferred (single)	0.068 (86)	0.035 (93)	0.073 (87)
	2.02	1.93	1.87

The number of populations was 2 and the number of individuals per population was 25.

ISSN: 1471-2105