I have been attempting to apply dihedral constraints to the phi/psi angles of a small TRP dipeptide minimization without success. From the output pdb it appears the second dihedral constraint is not applied to the CN atom of the Acetylayted end cap. This dihedral angle shifts to 150 degrees despite a very strong constraint. I suspect Rosetta might be selecting the first CN atom it finds which would be the methyl carbon on the NME cap rather than the carbonyl carbon on the ACE group of the same name.
My main goal here is to interpolate the 2D psi/phi potential energy surface for different canonical and non-canonical residues for comparison. For this I'll be applying dihedral constraints across the phi/psi grid, minimizing each point, then calculating a static score with any constraints removed.
I am using the linux_2020.08.61146_bundle with precompiled binaries.
Any help would be greatly appreciated.
PDB File (attached):
Constraints File: (This is a test case to try and fix dihedral angles)
Dihedral CN 1 N 1 CA 1 C 1 HARMONIC -3.14 0.05 Dihedral N 1 CA 1 C 1 NM 1 HARMONIC -3.14 0.05
-s pes.pdb -run:min_type lbfgs_armijo_nonmonotone_atol -run:min_tolerance 0.00000001 -constraints:cst_file cstfile -score:weights ref2015_cst -out:suffix _minwithcsts -optimization:default_max_cycles 1000000000