TABLE 1.
Outgroup | Probe sets | t1 | t2 | log2(t1/t2) | a | σ | p |
---|---|---|---|---|---|---|---|
No | No | 1.91 (1.72–2.08) | 0.36 (0.27–0.52) | 2.45 (1.83–2.75) | 0.50 (0.48–0.54) | 0.18 (0.12–0.18) | 0.00 (0.00–0.01) |
Yes | No | 2.02 (1.86–2.16) | 0.49 (0.43–0.56) | 2.00 (1.83–2.24) | 0.50 (0.47–0.53) | 0.12 (0.10–0.14) | 0.01 (0.01–0.01) |
No | Yes | 2.01 (1.39–2.40) | 0.50 (0.37–0.64) | 2.00 (1.49–2.51) | 0.43 (0.33–0.49) | 0.30 (0.22–0.35) | 0.00 (0.00–0.00) |
True parameters | 2.00 | 0.50 | 2.00 | 0.50 | 0.10 | 0.01 |
Shown are medians and 25 and 75% quartiles for estimates of branch lengths t1, t2, and log2(t1/t2) and shape parameters a, σ, and outlier proportion p over 20 simulated data sets with parameters close to values inferred from actual data (see Table 2). Inferences were made from data without (two-species tree) or with a single outgroup (three-species tree) and using either 1000 probes or 1000 probe sets, each consisting of a geometrically distributed number (mean 6) of identical probes. Full simulation results are presented in supplemental Tables 1–3.