I'd like to change the rosetta scoring function in remodel to favour models that have less hydrophobic area. Main reason is low computing power for the task at hand.
Changing fa_sol is definitely the option? Should i change the lk_ball_wtd as well?
Second thing is the assignment of secondary structure. I have clear helix propensity calculations at hand and rosetta seems to put out lots of models with wrong secondary structure even though i assigned H and L in blueprint per calculations. Should i reduce the rama_prepro or smth else?
Thanks for your time!
For design steps, I'd recommend using the aa_composition scoring term. It allows you define "composition constraints", each of which is a desired amino acid composition in a pose or sub-region of a pose (e.g. "I want no more than 10% alanine" or "I want 25% of residues to be valine, leucine, or isoleucine" or "I want no less than 75% polar residues"), and then to add a nonlinearly-ramping penalty for deviation from this composition. The steps are:
- Add your composition constraints (which can be done at the commandline or with the AddCompositionConstraintMover, https://www.rosettacommons.org/docs/latest/scripting_documentation/RosettaScripts/Movers/movers_pages/AddCompositionConstraintMover).
- Create a scorefunction with a nonzero weight for the aa_composition scoreterm.
- Pass this scorefunction to the mover that you use for design (presumably Remodel, in your case).
In your case, you'd presumably say "I want no more than x% hydrophobic residue types in the region that I'm designing".
Documentation is here: https://www.rosettacommons.org/docs/latest/rosetta_basics/scoring/AACompositionEnergy