. 2008 Mar;178(3):1817–1829. doi: 10.1534/genetics.107.081281

TABLE 7.

Results for the Singh and Rhomberg (1987) data set

	F_st				P-value			P(δ_i = 1 \| data):
Locus	Article	Original	Reparameterized	Variable selection	Article	Original	Reparameterized	Variable selection
G6-PD	0.47^a	0.50^a	0.50^a	0.31	0.00^a	0.01^a	0.01^a	0.12
ADH	0.45^a	0.49^a	0.49^a	0.30	0.01^a	0.01^a	0.01^a	0.09
EST-6	0.18^a	0.15	0.15	0.24	0.95^a	0.94	0.94	0.01
PT-9	0.43^a	0.41^a	0.41^a	0.27	0.05^a	0.04^a	0.04^a	0.02
PT-15^b	0.52^a	0.53^a	0.53^a	0.34^a	0.00^a	0.00^a	0.00^a	0.20^a
XDH	0.13^a	0.14^a	0.14^a	0.23	0.98^a	0.98^a	0.98^a	0.03
a-FUC	0.14^a	0.14^a	0.14^a	0.24	0.97^a	0.97^a	0.97^a	0.02
LAP-6	0.47^a	0.44^a	0.44^a	0.28	0.03^a	0.03^a	0.03^a	0.04
ACON-1	0.15^a	0.12^a	0.12^a	0.23	0.99^a	0.98^a	0.98^a	0.04
a-GLU-4	0.17^a	0.15^a	0.15^a	0.24	0.96^a	0.96^a	0.96^a	0.02

Estimated F_st-values for all methods, corresponding Bayesian P-values P(α_i < 0 | data) for the original and the reparameterized algorithm, and corresponding posterior probabilities P(δ_i = 1 | data) for the reparameterized algorithm including Bayesian variable selection for loci detected being subject to selection by one of the methods are shown. Article, Beaumont and Balding (2004) results; Original, original algorithm; Reparameterized, reparameterized algorithm; Variable selection, reparameterized algorithm including Bayesian variable selection.

The locus is classified subject to selection by the corresponding method.

All methods classify the corresponding locus subject to selection.