# Re: detecting vanishing atoms in zero.fep

Date: Wed Oct 31 2018 - 09:06:15 CDT

Please continue all conversations on namd-l. These messages are archived so
that they can be searched in the future.

How you assign groups is entirely dependent on: 1) what relative free
energy you wish to compute and 2) how you choose to construct your
topology. There is no single correct answer to this problem.

suggest writing out a thermodynamic cycle (or multiple!) and then choosing
endpoints that can be reasonably simulated. You can then assign alchemical
groups to match these endpoints at lambda = 0/1.

BKR

On Tue, Oct 30, 2018 at 6:50 PM soroush ziaei <soroush.ziaei7337_at_gmail.com>
wrote:

> dear brian,
> for a better prespective, we have a structure that contains three aromatic
> ring, so can i divide this three groups into -1,1,0 ?
> or in another words we have a lot of structures that has a symmetry. for
> this item can i choose number <1> for one side?
>
> On Mon, Oct 29, 2018 at 11:51 AM soroush ziaei <
> soroush.ziaei7337_at_gmail.com> wrote:
>
>> this problem under consideration! what do you mean brian?
>> meanwhile in this .rtf file (topology), can i put -1 & 1 for CG2R61
>> (C2-C14) and HGR61 respectively?
>>
>> # CGENFF.rtf
>>
>> RESI ***** 0.000 ! param penalty= 78.400 ; charge penalty= 25.148
>> GROUP ! CHARGE CH_PENALTY
>> ATOM N1 NG2R51 -0.169 ! 25.148
>> ATOM N2 NG2R50 -0.559 ! 5.159
>> ATOM N3 NG2R50 -0.662 ! 0.474
>> ATOM N4 NG1T1 -0.463 ! 0.000
>> ATOM N5 NG1T1 -0.463 ! 0.000
>> ATOM C1 CG311 0.289 ! 24.191
>> ATOM C2 CG2R61 -0.005 ! 14.165
>> ATOM C3 CG2R61 -0.005 ! 14.165
>> ATOM C4 CG2R61 -0.115 ! 6.325
>> ATOM C5 CG2R61 -0.115 ! 6.325
>> ATOM C6 CG2R61 -0.115 ! 6.325
>> ATOM C7 CG2R61 -0.115 ! 6.325
>> ATOM C8 CG2R53 0.266 ! 5.687
>> ATOM C9 CG2R61 -0.104 ! 0.000
>> ATOM C10 CG2R61 -0.104 ! 0.000
>> ATOM C11 CG2R61 -0.104 ! 0.000
>> ATOM C12 CG2R61 -0.104 ! 0.000
>> ATOM C13 CG2R61 0.095 ! 0.000
>> ATOM C14 CG2R61 0.095 ! 0.000
>> ATOM C15 CG2R53 0.468 ! 2.500
>> ATOM C16 CG1N1 0.356 ! 0.000
>> ATOM C17 CG1N1 0.356 ! 0.000
>> ATOM H1 HGA1 0.090 ! 0.601
>> ATOM H2 HGR61 0.115 ! 0.000
>> ATOM H3 HGR61 0.115 ! 0.000
>> ATOM H4 HGR61 0.115 ! 0.000
>> ATOM H5 HGR61 0.115 ! 0.000
>> ATOM H6 HGR52 0.157 ! 0.684
>> ATOM H7 HGR61 0.115 ! 0.000
>> ATOM H8 HGR61 0.115 ! 0.000
>> ATOM H9 HGR61 0.115 ! 0.000
>> ATOM H10 HGR61 0.115 ! 0.000
>> ATOM H11 HGR52 0.110 ! 0.000
>>
>> On Sun, Oct 28, 2018 at 7:09 PM Brian Radak <brian.radak_at_gmail.com>
>> wrote:
>>
>>> Your question seems entirely dependent on the problem under
>>> consideration.
>>>
>>> Atoms with a flag of 1 get scaled as a function of alchLambda while
>>> those with a flag of -1 get scaled as a function of 1-alchLambda. If you
>>> swap the groups you just change the sign of the free energy that is
>>> computed, so this is somewhat arbitrary and/or a matter of bookkeeping.
>>>
>>> HTH,
>>> BKR
>>>
>>> On Sat, Oct 27, 2018, 4:39 PM soroush ziaei <soroush.ziaei7337_at_gmail.com>
>>> wrote:
>>>
>>>> hi all,
>>>> i want to know how can i detect which atom should be -1 in fep file in
>>>> order to obtain correct value for delta<g>?
>>>> or please describe some method to generate topology file for a unique
>>>> structure, such as topology file that exist in tutorial (zero.top).
>>>>
>>>> best regards,
>>>> soroush
>>>>
>>>

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2019 - 23:20:17 CST