Table 3. Coalescent simulations with and without selection.

	Recombination parameter
	C=0	C=Rm	C=R
No selection
(A) Number of data sets	10 000	10 000	10 000
P (i⩾14)	0.120	0.064	0.013
(B) Number of data sets	1200	643	126
P (d<5∪d>30)	0.10	<0.01	<0.01
Selection coefficient, s=0.01
(C) Number of data sets	1000	1000	1000
P (i⩾14)	>0.90	>0.90	>0.85
(D) Number of data sets	976	974	954
P (d<5∪d>30)	0.05	<0.01	<0.001
Selection coefficient, s=0.1
(E) Number of data sets	1000	1000	1000
P (i⩾14)	>0.95	>0.95	>0.95
(F) Number of data sets	977	968	962
P (d<5∪d>30)	0.04	<0.01	<0.001

Probability of observing (A, C, E) a minimum of i identical sequences using a threshold value determined from empirical observations of the frequency of the Del200 haplogroup and (B, D, F) a major haplogroup that is highly divergent from all other alleles in the population whereby the pairwise distance, d, between the major allele and all other sequences in a data set is either <5 (representing variants within the Del200 haplogroup) or >30 (representing the 31 fixed differences between the Ins200 and Del200 haplogroups). All data sets were simulated using the parameters n=63 and S=80. For simulations with selection, N_e ranged from 10⁵ to 10⁷and the SF option with t=0 and f=0.3 was used (Ewing and Hermisson, 2010). In addition to a no recombination scenario, two estimates of the recombination parameter, C, were included: the minimum number of recombination events, R_m=5 (Hudson, 1987), and the estimator based on the variance of the average number of differences between pairs of sequences, R=21.7 (Hudson and Kaplan, 1985).