Table 5.
Summary of branch-site models implemented in PAML. All codons inferred to be under selection correspond to sites within each separate species alignment; for these tests, we used a critical posterior probability value of P > 0.90
| Species; Analysis | Genealogy | Branch-site hypothesis | −2ΔlnL (df = 1) | Parameter estimate (ω) | Positively selected sites (ω > 1; P > 0.90) |
|---|---|---|---|---|---|
| Pisaster ochraceus; Population-level | Maximum likelihood (ML) | Cordova | 8.341 | ω2 = 197 | 624K (P = 0.948) |
| Bamfield | 0 | ω2 = 1 | – | ||
| Lighthouse Park | 0 | ω2 = 1 | – | ||
| Hopkins | 5.111 | ω2 = 44.8 | – | ||
| La Jolla | 4.231 | ω2 = 77.4 | – | ||
| Bayesian | Cordova | 9.991 | ω2 = 455 | 624K (P = 0.971) | |
| Bamfield | 0 | ω2 = 1 | – | ||
| Lighthouse Park | 0 | ω2 = 1 | – | ||
| Hopkins | 5.031 | ω2 = 47.0 | – | ||
| La Jolla | 5.121 | ω2 = 107 | – | ||
| Pisaster brevispinus; Population-level | ML | Bamfield | 2.18 | ω2 = 188 | – |
| Port Moody | 0 | ω2 = 1 | – | ||
| SF Bay | 1.24 | ω2 = 1 | – | ||
| Bayesian | Bamfield | 0 | ω2 ≫ 1 | – | |
| Port Moody | 0 | ω2 = 1 | – | ||
| SF Bay | 0 | ω2 = 1 | – |
1
Significant (df = 1).