Skip to main content
letter
. 2019 May 1;3(3):299–312. doi: 10.1002/evl3.117

Table 2.

Estimates of the proportion of adaptive mutation (α) under various models of distributions of fitness effects

Data set Model Nb. parameters Log likelihood AIC α ωa
All genes Neutral 16 −1111.199 2254.398 0.253 0.031
All genes Gamma 17 −286.142 606.284 0.464 0.058
All genes GammaExpo 19 −229.425 496.850 0.352 0.044
All genes DisplacedGamma 18 −286.142 608.284 0.464 0.058
All genes ScaledBeta 18 −273.334 582.669 0.485 0.060
All genes BesselK 19 −294.387 626.774 0.514 0.064
Non‐effectors Neutral 16 −1104.260 2240.519 0.252 0.031
Non‐effectors Gamma 17 −286.821 607.643 0.463 0.057
Non‐effectors GammaExpo 19 −230.221 498.442 0.388 0.048
Non‐effectors DisplacedGamma 18 −286.821 609.643 0.463 0.057
Non‐effectors ScaledBeta 18 −273.687 583.375 0.458 0.057
Non‐effectors BesselK 19 −296.728 631.456 0.513 0.063
Effectors Neutral 16 −101.036 234.071 0.307 0.062
Effectors Gamma 17 −94.255 222.510 0.485 0.097
Effectors GammaExpo 19 −87.332 212.664 0.666 0.134
Effectors DisplacedGamma 18 −94.684 225.369 0.492 0.099
Effectors ScaledBeta 18 −86.845 209.689 0.600 0.120
Effectors BesselK 19 −87.416 212.832 0.507 0.102

AIC, Akaike's information criterion; α, proportion of adaptive substitutions; ωa, rate of adaptive substitutions. Values in bold indicate the best model fit for each gene set.