Dummy atoms

User discussions
1

GROMACS version: 2019.4
GROMACS modification: No

Hi everybody,

Willing to use dummy atoms (e.g. definition of O2 with O-dummy-O) in a polymer membrane, I have difficulties to minimize the structure (the forces on the dummy atom are too large). There must be something wrong with my construction.

I defined

  • in ffbonded.itp:
    OM OM 1 0.12440 347286.6
  • in ffnonbonded.itp :
    opls_972 OM 8 15.99940 0.000 A 3.58000e-01 9.81100e-01
    DUO DU 0 0.00000 0.8476 D 0.00000e+00 0.00000e+00
  • in O2M.top:
    –
    [ atoms ]
    ; nr type resnr residue atom cgnr charge mass typeB chargeB massB
    ; residue 1 O2M rtp O2M q 0.0
    1 opls_972 1 O2M O1 1 -0.4238 15.9994
    2 DUO 1 O2M DU 1 0.8476 0
    3 opls_972 1 O2M O2 1 -0.4238 15.9994 ; qtot 0

[ atoms ]
; nr type resnr residue atom cgnr charge mass typeB chargeB massB
; residue 1 O2M rtp O2M q 0.0
1 opls_972 1 O2M O1 1 -0.4238 15.9994
2 DUO 1 O2M DU 1 0.8476 0
3 opls_972 1 O2M O2 1 -0.4238 15.9994 ; qtot 0

[ bonds ]
; ai aj funct c0 c1 c2 c3
1 3 1

[ dummies2 ]
2 1 3 1 0.5

[ constraints ]
1 3 1 1.2

[ exclusions ]
1 2 3

I assume the problem is related to ‘dummies2’, ‘constraints’ and ‘exclusions’. What is it that I don’t understand?

Thanks,

Stefan

2

The right construction is the following:

–
[ moleculetype ]
; Name nrexcl
O2M 3

[ atoms ]
; nr type resnr residue atom cgnr charge mass typeB chargeB massB
; residue 1 O2M rtp O2M q 0.0
1 opls_972 1 O2M O1 1 -0.4238 15.9994
2 opls_972 1 O2M O2 1 -0.4238 15.9994 ; qtot 0
3 DUO 1 O2M DU 1 0.8476 0 ; qtot 0

[ bonds ]
; ai aj funct c0 c1 c2 c3
1 2 1

[ dummies2 ]
3 1 2 1 0.5

[ exclusions ]
1 2 3
3 1 2
2 1 3

Maybe a small note: ‘dummies2’ is depreciated - the keyword (with the same functions) ‘virtual_sites2’ is advocated nowadays.

The manual (Fig. 34 in https://manual.gromacs.org/documentation/2019-rc1/reference-manual/functions/interaction-methods.html#virtualsites ) gives numerator ‘2’ to a linear construction of a dummy atom. However, in ‘[ dummies2 ]’ (or ‘[ virtual_sites2 ]’) still ‘1’ is used (the last ‘1’ in 3 1 2 1 0.5 )