[gmx-developers] GROMACS-RAMD implementation using PULL code

Berk Hess hess at kth.se
Thu May 23 10:02:49 CEST 2019


Hi,

I presume you mean the pull external-potential interface, not AWH.

One solution would be using 3 direction pull coordinates: x, y and z. 
Maybe not very elegant, but it should work.

Cheers,

Berk

On 23/05/2019 09.56, Bernd Doser wrote:
> Dear Berk,
>
> Thank you for your fast answer.
>
> Using AWH with an new external force provider is exactly what I want to
> do. The only part which I can not do with AWH is changing the force
> direction. I think I have to overwrite the pull-direction directly
> on-the-fly. Therefore, I would have some special code for RAMD beside
> PULL/AWH, which would be affected by a ForceProvider migration.
>
> Cheers,
> Bernd
>
> On 5/23/19 09:37, Berk Hess wrote:
>> Hi,
>>
>> By now we might have all infrastructure in place to convert the pull
>> code to MDModule and ForceProvider, but we do not have a concrete time
>> plan for this.
>>
>> But for RAMD it would seem to me that you want to use exactly the same
>> coupling that the AWH module uses. You can register an external force
>> provider to for pull coordinates and supply forces to it yourself. Have
>> a look at awh.cpp and related files.
>>
>> Cheers,
>>
>> Berk
>>
>> On 23/05/2019 09.27, Bernd Doser wrote:
>>> Dear GROMACS developers,
>>>
>>> Currently I am working on a implementation of the 'Random Acceleration
>>> Molecular Dynamics' (RAMD) method into GROMACS. Please find details to
>>> RAMD at https://www.h-its.org/software/ramd/
>>>
>>> For the implementation into GROMACS we want to use the PULL code. I
>>> found some references in the code for intentions to migrate the PULL
>>> code into the new ForceProvider feature, which I would appreciate very
>>> much.
>>>
>>> Regarding to GROMACS-RAMD I would like to know, if there is a concrete
>>> schedule for the ForceProvider migration.
>>>
>>> Thank you in advance for any information.
>>>
>>> Best regards,
>>> Bernd Doser
>>>
>>>



More information about the gromacs.org_gmx-developers mailing list