[gmx-users] Rigid four atom molecule using LINCS

[Philip Loche]

Dear all,

Is there any way (possibly by use of virtual sites) to create a completely rigid molecule consisting of four atoms arranged in tetrahedral shape and simulate it using the LINCS constraints algorithm? 
Specifically we try to rebuild our H3O+ molecule below in a way that allows us to use LINCS rather than SHAKE, as this is not possible with the amount of constraints used in the .itp file below…

[ moleculetype ]
; molname       nrexcl
H3O             3

[ atoms ]
;   nr   type  resnr residue  atom   cgnr     charge       mass
     1     OI      1    H3O     OI      1    -1.4000   15.99940
     2     HI      1    H3O    HI1      1     0.8000    1.00800
     3     HI      1    H3O    HI2      1     0.8000    1.00800
     4     HI      1    H3O    HI3      1     0.8000    1.00800

[ constraints ]
; i     j       func    dist
1       2       1       0.098
1       3       1       0.098
1       4       1       0.098
2       3       1       0.1619
2       4       1       0.1619
3       4       1       0.1619

[ exclusions ]
; interactions between the first atom and all following atoms will be excluded
1       2       3       4
2       1       3       4
3       2       1       4
4       2       3       1

The .itp file is taken from:
Bonthuis, D. J., Mamatkulov, S. I., & Netz, R. R. (2016). Optimization of classical nonpolarizable force fields for OH- and H3O+. Journal of Chemical Physics, 144(10), 104503–164504. https://doi.org/10.1063/1.4942771

When using virtual sites, we of course want to retain the same rotational properties as in the case above, meaning an identical moment of inertia. This means the masses will have to be redefined appropriately I assume.

Many thanks in advance
Best regards

Philip