[gmx-users] mdrun with multi option

[Debdip Bhandary]

Thank you Mark.

Let me explain my scenario here - Umbrella sampling simulations with 15 
windows are run using -multi option to obtain PMFs. I have not noticed 
any issues with the output files (trr, pullf etc.) yet. I have carried 
out WHAM analysis using gmx wham option and obtained smooth PMFs. Please 
explain me in which way 'gmx mdrun -multi' option may lead to less 
reliable with output file and how I can be sure of it.

Thanks again.
regards,
Debdip

On 07.06.2017 12:39, Mark Abraham wrote:
> Hi,
>
> gmx mdrun -multidir is a good way to run a group of similar but otherwise
> uncoupled simulations - they remain uncoupled unless you choose something
> that couples them, such as replica exchange. gmx mdrun -multi is a bit less
> reliable with output files.
>
> Mark
>
> On Wed, Jun 7, 2017 at 12:34 PM Debdip Bhandary <
> debdip.bhandary at mv.uni-kl.de> wrote:
>
>> Dear Users,
>>
>> I have doubts regarding simulating multiple systems using mdrun '-multi'
>> option.
>>
>> I have seen that people have used this option for replica exchange (with
>> '-replex' option, of course) where information are exchanged between the
>> systems. In case of different systems (such as different windows in
>> Umbrella sampling simulations, liquid water systems at different
>> temperatures, etc.), as per my understanding (from user manual), there
>> is no exchange of information between the systems and simulations are
>> run without any influence of each other. Am I correct here? or is there
>> any issue when different systems are simulated using -multi/multidir
>> options? Any information regarding this would be appreciated.
>>
>> Thank you in advance.
>>
>> Best regards,
>> Debdip Bhandary