[gmx-users] Missing LJ interaction
fiedler at umich.edu
Tue Apr 3 16:19:09 CEST 2007
David van der Spoel wrote:
> Steve Fiedler wrote:
>> Thank you for looking over the input and for suggestions Mark.
>> Unfortunately, a comparison of the topologies with diff revealed no
>> inconsistencies. GROMACS can correctly recognize the presence of the
>> significant repulsion (e+10) between atoms 1 and 5 when the atoms
>> from both triatomic molecules are incorporated into a single molecule.
>> I have found one clue in that the missing repulsion in the
>> bimolecular system, can be returned by increasing rlist >= 0.77 nm in
>> grompp.mdp. This is unsatisfying, as the distance between atoms 1
>> and 5 is much shorter, 0.057 nm. As these coordinates were taken
>> from a calculation of membrane system just before a crash, I believe
>> it is this lack of intermolecular nonbonding repulsion responsible
>> for allowing atoms 1 and 5 to get this close in the first place.
>> Again, any insights are most welcome.
> I didn't follow the thread but it is the distance between charge
> groups that determines whether interactions are incorporated in the
> neighborlist. Check the manual about charge groups.
This issue has been resolved. Thank you Prof. van der Spoel for
providing direction. Below is the corrected topology file of the
original bimolecular system. From the mailing list archive, the
following messages were helpful:
[ defaults ]
; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ
1 2 no 0.125 0.5
[ atomtypes ]
;name mass charge ptype sig eps
C2 14.0300 0.300 A 0.39057 0.49371
C3 15.0300 0.400 A 0.39057 0.73220
O 16.0000 -0.694 A 0.30647 0.71128
[ nonbond_params ]
; i j func sig eps
C2 C2 1 0.39057 0.493712000
C2 C3 1 0.39057 0.601240800
C3 C3 1 0.39057 0.732200000
C2 O 1 0.34852 0.592579920
C3 O 1 0.34852 0.721656320
O O 1 0.30647 0.711280000
[ moleculetype ]
[ atoms ]
; nr type resnr residue atom cgnr charge mass
1 C2 1 1CHO C2 0 0.000 14.03
2 C3 1 1CHO C3 1 0.000 15.03
3 O 1 1CHO O 2 0.000 16.00
[ system ]
[ molecules ]
; name number
Steve Fiedler, Ph.D.
Department of Mechanical Engineering
The University of Michigan
2024 G.G. Brown
2350 Hayward St.
Ann Arbor, MI 48109-2125
More information about the gromacs.org_gmx-users