I am struggling with adjusting my NOE constraints and hope that someone with a little more experience can share some guidance.
My problems are:
1) I get local distortions, and am not sure how to best set the NOE constraints
2) Both constrained helices often move, but I only want one of them to move and the other one to stay in place.
I am modelling a GPCR, a transmembrane protein with a bundle of 7 alpha-helices. I am applying NOE distance constraints between helices to obtain an active from an inactive conformation. These look like this:
ASSIGN select resid @resida .and. type @typea end -
select resid @residb .and. type @typeb end -
RMIN @currentdistconstmin RMAX @currentdistconstma KMIN 50 KMAX 50 FMAX 150
To avoid local distortions within the alpha-helices I have also applied NOE constraints to the backbone hydrogen bonds. This constrain is stronger with higher force constants and FMAX:
ASSIGN select segid proa .and. resid @firstres .and. type N end -
select segid proa .and. resid @lowerres .and. type O end -
RMIN 2.7 RMAX 3.6 KMIN 100 KMAX 100 FMAX 300
To reduce the extent that my protein has to move at one time, I am applying my constraints gradually in 10 equal increments of 100 ps dynamics each. The differences between the actual distances and my target distances (delta Rs) are typically between 0.1-1 Angstrom. I am currently testing 300 ps for each increment to see if this can reduce the local distortions.
I am considering adding harmonic bestfit constraints to the constrained atoms in helices that should not move, like CONS HARM BESTFIT .... However, this might result in the rest of the helix moving and leaving the atom with the harmonic constraints misplaced? I could of course add harmonic constraints to more atoms in the helices that should not move, but I want to keep the harmonically constrained atoms at a minimum.
Does anyone with experience have any tips?