n | L | FN GT | FN HT | FP GT | FP HT |
|---|
50 | 10 | 30% | 9.9% | 0.67% | 0.29% |
100 | 10 | 26% | 13% | 1.5% | 0.65% |
200 | 10 | 32% | 28% | 3.4% | 0.75% |
50 | 5 | 30% | 2.9% | 0.83% | 0.19% |
100 | 5 | 15% | 5.6% | 3.3% | 0.85% |
200 | 5 | 19% | 13% | 7.1% | 1.6% |
50 | 2 | 29% | 2.6% | 5.4% | 4.8% |
100 | 2 | 18% | 2.4% | 16% | 14% |
200 | 2 | 14% | 5.5% | 39% | 34% |
- A constant size population of N = 10000 diploids of L centimorgans containing s = 400 SNPs was simulated for T = 100 generations using a Wright-Fisher model, then a family of size n was sampled and simulated as a small population for t = 5 generations. The last four columns measure, after performing IBD detection, the rate of false positives using genotype in the FN GT column, and using phased genotype in the FN HT column, and the rate of false positives using genotype in the FP GT column, and using phased genotype in the FP HT column. Parameter choices when using genotype were γ = 8, δ = 12, and λ = μ = 1/ 2, and when using phased genotype were γ = 8, δ = 32, and λ = μ = 1 = 256. Results were averaged over 10 different simulations per choice of parameters.
