Potential | Model |
y
|
x
|
AIC
|
w(AIC)
|
R
|
pR
|
---|
pfANM | Stress |
ω
|
MLmS
| 191424 | 1.00 | -0.55 | -0.32 |
| Flexibility |
ω
|
MSF
| 199645 | 0.00 | 0.47 | -0.04 |
ANM | Stress |
ω
|
MLmS
| 194589 | 1.00 | -0.52 | -0.40 |
| Flexibility |
ω
|
MSF
| 207993 | 0.00 | 0.36 | -0.02 |
- NOTE: Results of global fits for all sites of the dataset. Potential is the ENM potential. Model is either the Stress Model or Flexibility Model. y is in all cases the site-specific rate of evolution inferred using Rate4Site, x is the independent variable of each model. AIC is the Akaike Information Criterion, that quantifies the goodness of fit of a model (the lower the better). w(AIC) is the AIC weight of evidence for each compared model (same ENM). R is Pearson’s correlation coefficient between variables y and x. pR is the partial correlation coefficient when controlling for the x variable of the contrasting model (same ENM). All variables were z-score normalized for each protein before fitting.