I was trying to mix ligand docking with protein design of a symmetric dimer by using the SetupNCS mover. I still don't know why the mover is not being used...the resultant pdbs were modified at the sequence level but both chains are different.
thanks in advance
The script used is:
<SetupNCS name=ncs bb=1 chi=1 symmetric_sequence=1>
<NCSend source="115B-227B" target="1A-113A"/>
<NCSend source="1A-113A" target="115B-227B"/>
<Translate name=translate chain=X distribution=uniform angstroms=0.25 cycles=50/>
<Rotate name=rotate chain=X distribution=uniform degrees=360 cycles=500/>
<EnzRepackMinimize name=desmin design=1 repack_only=0 scorefxn_minimize=myscore scorefxn_repack=soft_rep minimize_rb=1 minimize_sc=1 minimize_bb=0 cycles=2 minimize_lig=1 min_in_stages=0 backrub=0 task_operations=liginter/>
I realize now that it will symmetrize my dimer, but it is not what I was thinking...do you have any suggestion?
Unlike normal symmetry which enforces an obligate symmetry in the modification itself, the non-crystallographic symmetry mover enforces the symmetry via energetic restraints on the deviations between the residues. This means that in order to acutally apply the symmetry between the subunits, you need to be using a scorefunction which applies the energetic penalty while moving/designing.
Unfortunately, the default scorefunction does not have the appropriate energy terms turned on. What you'll want to do is to turn on (set to 1.0 or maybe more) the "res_type_constraint" weight in your "myscore" and "soft_rep" energy functions. (Add the dihedral_constraint term too, if you want the structure to be maintained as symmetric as well as the identity.)