Table 4.
Bayes factor comparisons between model M0 and M1 for each ERV gene showing also the number of generations for each Bayesian inference.
| Gene | Model comparison1 | Number of generations | 2*loge B10§ | ||
|---|---|---|---|---|---|
| M0 | M1 | M0 | M1 | ||
| pol | *1-HKY+Γ | +3-HKY+Γ | 3,000,000 | 6,000,000 | 0.049 |
| gag | *1-K80+Γ | +3-K80+Γ | 8,000,000 | 13,000,000 | 0.033 |
| env (class A) | *1-HKY+Γ | +3-HKY+Γ | 3,000,000 | 10,000,000 | -0.061 |
| env (class B) | *1-K80 | +3-K80 | 3,000,000 | 3,000,000 | 0.016 |
| env (class E) | *1-HKY+Γ | *1-GTR+Γ | 3,000,000 | 3,000,000 | 0.0025 |
| *1-HKY+Γ | +3-HKY+Γ | 3,000,000 | 3,000,000 | 0.0014 | |
1. HKY = Hasegawa-Kishino-Yano model [78]; K80 = Kimura 2-parameter [79]; Γ = gamma distribution.
* 1 = one model for the whole gene.
+ 3 = partition of the data per codon position.
§ Values represent twice the loge of the Bayes factor (B10) in the comparison of models M0 and M1 (2*loge B10). Positive values below 2 indicate no evidence against M0 [71,80]