[gmx-developers] FW: Questions - cutoff inthebondingpotentialandediting the replica exchange function

Mark Abraham Mark.Abraham at anu.edu.au
Sun Apr 1 10:52:05 CEST 2012 

On 1/04/2012 6:41 PM, Asaf Farhi wrote:
> Dear Mark
>
> Thank you very much for the reply!
> Do you know how can I change the potential so it'll be the second one (or the forces)?
> If it can't be done by the user is anyone willing to help us with that?

Use a normal cut-off, and for every interaction past the cut-off, add 
the constant term to the energy. Forces are the same in either one of 
your cases, so I don't understand your reference to forces above. Now 
how are you going to deal with periodicity? And how does this even 
matter for your proposed exchange test?

Mark

>
> Thanks again,
> Best regards,
> Asaf
> ________________________________________
> From: gmx-developers-bounces at gromacs.org [gmx-developers-bounces at gromacs.org] on behalf of Mark Abraham [Mark.Abraham at anu.edu.au]
> Sent: Sunday, April 01, 2012 1:57 AM
> To: Discussion list for GROMACS development
> Subject: Re: [gmx-developers] FW: Questions - cutoff inthebondingpotentialandediting the replica exchange function
>
> On 31/03/2012 7:26 PM, Asaf Farhi wrote:
>> Dear Mark
>> Thank you very much for the reply!
>> I read 6.7, about coulombtype =Cut-off, and wanted to ask if it means:
>>
>> if (V<Ecutoff)
>>      V=0;
>> or
>> if (V>Ecutoff)
>>      V=Ecutoff;
>>
>> I need the second one and I didn't manage to understand which one is it.
> The first one is implemented, though it is usually phrased as (r>
> r_cutoff) =>  (V=0), since a Coulomb potential has V->0 as r->inf. You
> have to use some kind of a cut-off, as the number of interactions you
> have to compute grows with dvolume/d(r_cutoff), i.e. as (r_cutoff)^2.
>
> For normal MD the value of V at any point is irrelevant, because only
> the forces appear in the equations of motion. The value of (some
> components of) V might appear in the RE equations, but only if they
> don't cancel out for candidate exchange pairs in the subtraction of the
> exchange test does the value of V past the cutoff become relevant.
>
> Mark
>
>> Thanks in advance,
>> Best regards,
>> Asaf
>> ________________________________________
>> From: gmx-developers-bounces at gromacs.org [gmx-developers-bounces at gromacs.org] on behalf of Mark Abraham [Mark.Abraham at anu.edu.au]
>> Sent: Saturday, March 31, 2012 1:36 AM
>> To: Discussion list for GROMACS development
>> Subject: Re: [gmx-developers] FW: Questions - cutoff in thebondingpotentialandediting the replica exchange function
>>
>> On 31/03/2012 4:49 AM, Asaf Farhi wrote:
>>> Dear Cristoph
>>>
>>> Thank you very much for the reply.
>>> I looked at 4.2.13 and it seems that the bonds and angles terms can be edited. Do you know how can one edit the potentials between unbonded atoms (VDW and electrostatic)
>> Tabulated nonbonded interactions. Manual 6.7.
>>
>>> I need the same cutoff for all the interactions.
>> Easily done already.
>>
>> Mark
>>
>>> About the replica exchange I need it to be H-REMD (constraints and bonding potentials maybe don't need to be multiplied by lambda).
>>> I still haven't been to the office (I was in a trip).
>>> Does anyone know when will the new version of GROMACS be released? Alternatively is someone willing to cooperate with us on that?
>>> I'll send all the material if anyone considers it.
>>>
>>> Thanks in advance,
>>> Have a good weekend,
>>> Asaf
>>>
>>>
>>> ________________________________________
>>> From: gmx-developers-bounces at gromacs.org [gmx-developers-bounces at gromacs.org] on behalf of Christoph Junghans [junghans at votca.org]
>>> Sent: Thursday, March 29, 2012 12:45 AM
>>> To: Discussion list for GROMACS development
>>> Subject: Re: [gmx-developers] FW: Questions - cutoff in thebondingpotentialand editing the replica exchange function
>>>
>>> Am 28. März 2012 16:19 schrieb Asaf Farhi<asaf.farhi at weizmann.ac.il>:
>>>> Dear Christoph
>>>>
>>>> Thank you very much for the reply.
>>>> Cutoff in the energy per pair would be good. How is tabulating done?
>>> See manual section 4.2.13 and the gmx-users mailings list.
>>>
>>>> About the replica exchange. Most of the chances I'll need to change it. It's something similar to H-REMD and I can send you the details as soon as I get to the office.
>>> The place to look for the exchange rates is the function
>>> get_replica_exchange inside src/kernel/repl_ex.c
>>>
>>> Christoph
>>>> Thanks again,
>>>> Best regards,
>>>> Asaf
>>>>
>>>> ________________________________________
>>>> From: gmx-developers-bounces at gromacs.org [gmx-developers-bounces at gromacs.org] on behalf of Christoph Junghans [junghans at votca.org]
>>>> Sent: Wednesday, March 28, 2012 8:00 PM
>>>> To: Discussion list for GROMACS development
>>>> Subject: Re: [gmx-developers] FW: Questions - cutoff in the bondingpotentialand editing the replica exchange function
>>>>
>>>> Hi Asaf,
>>>>
>>>> Am 28. März 2012 10:44 schrieb Asaf Farhi<asaf.farhi at weizmann.ac.il>:
>>>>> Dear Professor Shirts
>>>>>
>>>>> Thank you very much for the reply.
>>>>> I thought about it more. It will also be good for us to use cutoff in the energy. This cutoff is for the maximal energy (for the nonbonded interactions).
>>>>> Editing the RE function might be unnecessary (Another way of implementation).
>>>>> In other words what is needed is to use cutoff for the energy so it won't be larger than a certain value (total VDW and electrostatic) and this value should be constant for all pairs.
>>>>> e.g
>>>>> if (V>cutoff)
>>>>>     V=cutoff;
>>>> I am not sure, do you want to cutoff the total bonded energy or just
>>>> the energy per pair (interaction function)? The latter can easily be
>>>> done by tabulating the bonded interaction.
>>>>
>>>> And what do you want to change about the replica exchange function?
>>>>
>>>> Cheers,
>>>>
>>>> Christoph
>>>>
>>>>> Using cutoff in the bonding energy can also be used but it seems complicated and I'm not sure necessary.
>>>>> Another small thing is if you know if you can exclude certain atoms from the non bonded interactions?
>>>>> Could you help us with that?
>>>>>
>>>>> Many thanks,
>>>>> best regards,
>>>>> Asaf
>>>>>
>>>>>
>>>>> ________________________________________
>>>>> From: gmx-developers-bounces at gromacs.org [gmx-developers-bounces at gromacs.org] on behalf of Shirts, Michael (mrs5pt) [mrs5pt at eservices.virginia.edu]
>>>>> Sent: Tuesday, March 27, 2012 3:50 PM
>>>>> To: Discussion list for GROMACS development
>>>>> Subject: Re: [gmx-developers] FW: Questions - cutoff in the bondingpotential and editing the replica exchange function
>>>>>
>>>>> Hi, Asaf-
>>>>>
>>>>> I think you need to be more specific to get any useful response.
>>>>>
>>>>>> One is about how to add cutoff to the bonding energy (I guess the pairing
>>>>>> terms can be done without editing the code)
>>>>> It's not clear what the purpose of adding a cutoff to the bonded term would
>>>>> be.
>>>>>
>>>>>> and the other one is if anyone
>>>>>> wants to help us in editing the replica exchange function (cooperation)?
>>>>> You need to specify what you are editing it to do!
>>>>>
>>>>> Best,
>>>>> ~~~~~~~~~~~~
>>>>> Michael Shirts
>>>>> Assistant Professor
>>>>> Department of Chemical Engineering
>>>>> University of Virginia
>>>>> michael.shirts at virginia.edu
>>>>> (434)-243-1821
>>>>>
>>>>>
>>>>> --
>>>>> gmx-developers mailing list
>>>>> gmx-developers at gromacs.org
>>>>> http://lists.gromacs.org/mailman/listinfo/gmx-developers
>>>>> Please don't post (un)subscribe requests to the list. Use the
>>>>> www interface or send it to gmx-developers-request at gromacs.org.
>>>>> --
>>>>> gmx-developers mailing list
>>>>> gmx-developers at gromacs.org
>>>>> http://lists.gromacs.org/mailman/listinfo/gmx-developers
>>>>> Please don't post (un)subscribe requests to the list. Use the
>>>>> www interface or send it to gmx-developers-request at gromacs.org.
>>>> --
>>>> Christoph Junghans
>>>> Web: http://www.compphys.de
>>>> --
>>>> gmx-developers mailing list
>>>> gmx-developers at gromacs.org
>>>> http://lists.gromacs.org/mailman/listinfo/gmx-developers
>>>> Please don't post (un)subscribe requests to the list. Use the
>>>> www interface or send it to gmx-developers-request at gromacs.org.
>>>> --
>>>> gmx-developers mailing list
>>>> gmx-developers at gromacs.org
>>>> http://lists.gromacs.org/mailman/listinfo/gmx-developers
>>>> Please don't post (un)subscribe requests to the list. Use the
>>>> www interface or send it to gmx-developers-request at gromacs.org.
>>> --
>>> Christoph Junghans
>>> Web: http://www.compphys.de
>>> --
>>> gmx-developers mailing list
>>> gmx-developers at gromacs.org
>>> http://lists.gromacs.org/mailman/listinfo/gmx-developers
>>> Please don't post (un)subscribe requests to the list. Use the
>>> www interface or send it to gmx-developers-request at gromacs.org.
>> --
>> gmx-developers mailing list
>> gmx-developers at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-developers
>> Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-developers-request at gromacs.org.
> --
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-developers-request at gromacs.org.

More information about the gromacs.org_gmx-developers mailing list