Table 2

Intensity of purifying selection (1-f) and levels of adaptive evolution (α)

Model name and description

f

α

ΔAIC

wAIC


M0: Strict selective neutrality

1

0

321

~ 0


M1: Homogeneous purifying selection +

No adaptive evolution

0.13

0

2.16

0.06


M2: Purifying selection with variable intensity +

No adaptive evolution

f1 = 0.11

f2 = 0.17

f3 = 0.08

0

1.16

0.16


M3 Homogeneous purifying selection + homogeneous levels of adaptive evolution

0.11

0.36

0.79

0.24


M4 Purifying selection with variable intensity + homogeneous levels of adaptive evolution

f1 = 0.10

f2 = 0.14

f3 = 0.07

0.35

0

0.52


M5: Purifying selection with variable intensity + varying levels of adaptive evolution

f1 = 0.09

f2 = 0.14

f3 = 0.09

α1 = 0.4

α2 = 0.37

α3 = -0.04

3.48

0.02


Model average (over M0 to M5)

f1 = 0.08

f2 = 0.11

f3 = 0.06

α1 = 0.28

α2 = 0.028

α3 = 0.027

N/A

N/A


1-f quantifies the intensity of purifying selection through the fraction of new amino acid changing mutation under strong purifying selection. α is the fraction of divergence attributable to adaptive evolution (driven by positive selection on amino acid changing mutations).

The best model (M4) is highlighted in bold and all differences in AIC among models ΔAIC) are computed relative to M4. WAIC are the statistical weights of each model in the model averaging procedure we used.

Subscripts in f1, f2, f3, α1, α2, α3 refer to the gene fragment categories. 1: control fragments, 2: flowering gene fragments, 3: symbiotic gene fragments. N/A: not applicable.

De Mita et al. BMC Evolutionary Biology 2011 11:229   doi:10.1186/1471-2148-11-229

Open Data