Hello.
I have some questions on the env energy term:
The values in the EnvPairPotential::env_log_ table should be the –ln[P] values defined for the env term in Rohl et al. 2004. Methods Enzymol. 383:66-93 (Table I). If the P's are sheer probabilities, then the –ln[P] values should all be positive or zero. However, looking at the input file from which the –ln[P] values are loaded (database/scoring/score_functions/EnvPairPotential/env_log.txt), it can be seen that many of them are negative. Why?
Is it because of interpolation? Interpolation is helpful if one wants to minimize env energies using derivatives, but derivatives are not defined for this term (the EnvEnergy class does not override OneBodyEnergy::eval_residue_derivatives(){}). This leads to the second question: Why is it that such an important energy term to centroid mode as the env term is left unminimized during derivative-based minimization?
Finally, the code in EnvEnergy::residue_energy() indicates that the env energy is always multiplied by 2.019, regardless of the amino acid type or neighbor count. What is the meaning of this multiplier?
Thanks,
Andrea.