[gmx-users] Including water in for restraining

[chris.neale at utoronto.ca]

you are correct that you must include the water as part of the same  
molecule (as you appear to have already done for Zn). I assume that  
protein_ion_A.itp contains only one [ moleculetypes ] section that has  
the protein and also the zinc. Take a look at how Zn has been included  
there. Repeat this procedure to include the water that you want to be  
bonded. Then modify your coordinate file accordingly. You can get  
water parameters from gromacs-3.3.1/share/gromacs/top/tip3p.itp (or  
sub the name of your favourite water model).

Note that this is intended as my best idea about how to go about doing  
exactly what you have asked. I have no idea how this is going to  
affect things like a temperature group that contains a protein, a  
zinc, and a single water if you have multiple temperature coupling  
groups. Also I am not sure if settle is going to work properly.

There is no warranty associated with this next portion, I have never  
tried it. I have included this just to be explicit about what I am  
suggesting. Lets say you use tip4p and oplsaa:

under [ atoms ] add:
2901       opls_113        1       SOL      OW     1       0.0
2902       opls_114        1       SOL     HW1     1       0.52
2903       opls_114        1       SOL     HW2     1       0.52
2904       opls_115        1       SOL      MW     1      -1.04

under [ bonds ] add:
#ifdef FLEXIBLE
; i     j       funct   length  force.c.
2901       2902       2901       0.09572 502416.0 0.09572        502416.0
2901       2903       2901       0.09572 502416.0 0.09572        502416.0
#endif

under [ angles ] add:
#ifdef FLEXIBLE
; i     j       k       funct   angle   force.c.
2902       2901       2903       1       104.52  628.02  104.52  628.02
#else
[ settles ]
; OW    funct   doh        dhh
2901       1       0.09572    0.15139
#endif

Then also add:
[ exclusions ]
2901       2902       2903       2904
2902       2901       2903       2904
2903       2901       2902       2904
2904       2901       2902       2903

[ dummies3 ]
; Dummy from                    funct   a               b
2904       2901       2902       2903       1       0.128012065      
0.128012065


Then go back and add any bonds that you want between this water and  
the protein or zinc.

It is unclear to me if the [ exclusions ] section is necessary or even  
advisable.