[gmx-users] Re: EDO and TRS ligand

[Mark Abraham]

On 14/04/2011 7:06 PM, Tsjerk Wassenaar wrote:
>
>> I am using the wavefunction spartan 8 program to calculate charges.
>> revised_EDO.itp
>> ...
>> [ moleculetype ]
>> ; Name nrexcl
>> EDO      3
>>
>> [ atoms ]
>> ;   nr      type  resnr resid  atom  cgnr   charge     mass
>>       1        OA     1  EDO     OAB     1   -0.530  15.9994
>>       2         H       1  EDO     HAE     1    0.332   1.0080
>>       3       CH2     1  EDO     CAA     1    0.198  14.0270
>>       4       CH2     1  EDO     CAC     2    0.198  14.0270
>>       5        OA      1  EDO     OAD    2    -0.530   15.9994
>>       6         H       1  EDO     HAF     2     0.332  1.0080
>> I didnt change dihedral angel, bond .... do I have to change them? I didnt
>> arranged for (COH)3(NH3+)C to have charges consistent with its symmetry (TRS
>> ligand). I have different charge for CH2 groups, H, O. That is, There is not
>> symmetry. What should I do?

An solvated TRS will be best modeled with the same set of partial 
charges on each symmetric side group (e.g. COH). So every oxygen atom 
should have the same partial charge. To do this, generate a structure of 
this compound that has the three-fold symmetry so that whatever 
technique you use to generate charges will do so symmetrically. Then, 
test them extensively against experimental data.

When bound as a ligand, those charges might differ between 
ostensibly-equivalent atoms, but trying to parametrize the charges in 
the context of the binding site is outside your scope at this time.

Mark