I evaluated same structure by Rosetta++(2.3.1) and PyRosetta 1.1(the latest miniRosetta), but they provided different total scores. Are there some changes for standard full-atom scoring function since the latest update?
The hub for Rosetta modeling software
There are many changes between the 2.3 and 3.x versions of the score function. We believe they are all improvements.
Some of the specific changes:
A) count-pair has been modified to be more internally consistent: 3-bond atoms have 0 interaction, 4 bond has .5, and 5+ has full.
B) some residue pairs used to be spuriously unscored; that's fixed
C) the whole etable (fa_rep, fa_atr, solvation) has been smoothed and its "range" lengthened a bit.