TABLE 1

Adjusted negative log-likelihood scores of the two models for complementation

Double heterozygote p3Heterozygote
p1
Heterozygote
p2
Model 1:
additive
Model 2:
synergistic
unc-13(e51)/+; unc-64(js115)/+unc-13(e51)/+unc-64(js115)/+14.9815.92
unc-13(n2813)/+; unc-64(e246)/+unc-13(n2813)/+unc-64(e246)/+118.7897.76
unc-13(n2813)/+; unc-64(js115)/+unc-13(n2813)/+unc-64(js115)/+117.7886.04
unc-13(e51)/+; unc-64(e246)/+unc-13(e51)/+unc-64(e246)/+41.4340.72
unc-64(e246)/+; snb-1(md247)/+unc-64(e246)/+snb-1(md247)/+47.5340.26
unc-13(n2813)/+; snb-1(md247)/+unc-13(n2813)/+snb-1(md247)/+50.1848.52
unc-13(n2813)/+; unc-18(e81)/+unc-13(n2813)/+unc-18(e81)/+60.0154.16
unc-13(n2813)/+; dpy-23(e840)/+unc-13(n2813)/+dpy-23(e840)/+48.5739.22
unc-13(n2813)/+; unc-104(e1265)/+unc-13(n2813)/+unc-104(e1265)/+50.1147.56
unc-13(n2813)/+; cha-1(pr1152)/+unc-13(n2813)/+cha-1(pr1152)/+30.0831.07
unc-13(n2813)/+; unc-17(e245)/+unc-13(n2813)/+unc-17(e245)/+30.2630.63
unc-13(n2813)/+; unc-29(e1072)/+unc-13(n2813)/+unc-29(e1072)/+45.1733.31
unc-13(n2813)/+; syd-2(ju37)/+unc-13(n2813)/+syd-2(ju37)/+27.9028.91
  • The additive model plays the role of a null model, as it has only two parameters, while the synergistic model has three. We compare the likelihood scores of the two models, adjusted for the number of parameters by the Akaike information criterion (AIC). This penalizes the higher parameter model, because an increase in parameters will always bring a better fit. When the synergistic model has a better AIC than the additive model, we reject additivity. The underlined score represents the accepted model.