. Author manuscript; available in PMC: 2008 Dec 7.

Published in final edited form as: Genet Epidemiol. 2005 Dec;29(4):353–364. doi: 10.1002/gepi.20092

Table 3.

Comparisons between FSA and H-clust for their performances on six simulated datasets based on various coalescent models^a.

			RSQ		PDE

Model	#SNP	N	FSA(0.5)	H-clust	FSA(0.5)	H-clust
Standard	45	11	0.976	0.944	0.949	0.864
Common(f > 0.1)	34	11	1.000	1.000	1.000	1.000
Island(mixed)	49	12	0.986	0.960	0.983	0.898
Island(single)	53	11	0.992	0.982	0.989	0.932
Expansion(t=500)	59	11	0.804	0.439	0.881	0.475
Expansion(t=5000)	51	11	0.876	0.924	0.915	0.898

The six coalescent models under which the datasets were simulated are as follows.

Standard: constant-size random mating population; Common: same random mating model as the previous one, but only retaining SNPs having minor allele frequency f > 0.1; Island(mixed): two-island model with migration rate of 1.0 per generation (an equal number of haplotypes was sampled from each island); Island(single): same island model as the previous one but with all samples coming from the same island; Expansion(t=500): exponentially expanding population with fast expansion starting 500 generations ago; Expansion(t=5000): exponentially expanding population with slow expansion starting 5000 generations ago.