TABLE 4

Coverage probability of the confidence intervals and average bias of the point estimates as functions of number of sequences included in the sample

nTCoverageBias wkH
1064230.9720.00100104345225612.21
2065040.9730.0016392641059512.85
5065290.9817.97 × 10-4820383218513.31
10069140.969-0.00159848378329813.47
• n is the number of sequences in the sample; denotes the square root of the mean squared error averaged over all simulated genealogies; w is the mean width of the 95% confidence intervals. Other population parameters are fixed: ℓ = 50,000; μ = 2.5 × 10-8; N0 = 4 × 105; g = 0.001. The quantities T, bias, H, and k are averaged over 1000 simulated genealogies for a given set of parameters.