[gmx-users] Trajectory guided by density maps

[Christian Blau]

Hi Ramon,


This feature will be released in January ;) and is in a beta version, soon to be release candidate, so you are 
definitely an early adopter.

Most of the parameters are set by default in the .mdp file if you do not set them yourself, you'll see them also in

http://manual.gromacs.org/2020-beta3/user-guide/mdp-options.html


It is hard to give very general advice on the parameters, but some things to observe:

  - the force constant can have a very large range of values, try changing the order of magnitude first, if you don't 
have an idea, or try "adaptive force scaling"

  - the resolution parameter density-guided-simulation-gaussian-transform-spreading-width is important as well and 
interplays with the force-constant - if this value is high (low resolution), the forces will go down as well (there is 
not so much force needed to make a blob be similar to anything)

  - density-guided-simulation-nst may be crucial for performance, the blobbier the densities and / or previous value the 
larger you can set this value. 20 or so is a good first guess for lots of applications


For now you cannot apply forces from two density maps at the same time, though this is an obvious possible extension for 
a next release.

Since I just implemented the feature, I'll be interested in how things work out for you and what is yet unclear in 
documentation, etc..


Best,

Christian


On 2019-12-12 11:58, Ramon Guixà wrote:
> Hi Christian,
>
> thanks a lot for your answer. Good to know this makes sense to someone else
> too.
> Now I have two questions out of ignorance (never used this newly
> implemented feature before):
>
> 1) Can I apply two sets of forces (restraints) at the same time, namely one
> for each density map?
> 2) Would you have by any chance an example or a default set of values for
> the density-guided options? I am not really sure how to use this feature
> anyway...
>
> Thanks!
> R
>
> On Thu, Dec 12, 2019 at 11:29 AM Christian Blau <blau at kth.se> wrote:
>
>> Hi Ramon,
>>
>>
>> This is definitely a way to obtain a transition path.
>>
>> One things to consider here is that you deliberately "throw away" already
>> perfect information about your target
>> structure, so it might be harder to reach. On the other hand you might not
>> care to have an exact overlap with the
>> structures, so in this sense using the density might aid you in finding a
>> more realistic transition path that is closer
>> to the lowest free energy barrier. If you're interested in that, using
>> very low resolution maps might even be beneficial.
>>
>> I personally would be excited to hear about your findings.
>>
>> Best,
>>
>> Christian
>>
>>
>> On 2019-12-11 17:18, Ramon Guixà wrote:
>>> Hi there,
>>>
>>> I would like to use GROMACS to create a trajectory that goes from one
>>> crystal structure to another crystal structure, both obtained from the
>> same
>>> system using time resolved crystallography.
>>>
>>> Would it be feasible to use the newly implemented density-guided
>>> simulations parameters to guide this simulation by first fitting the
>>> initial crystal structure its density map and then switching off these
>>> restraints while switching on the density map restraints of the second
>>> crystal structure? I figured this would in principle create a trajectory
>>> that better reflect the transition from one state to the other, but I am
>>> not even sure if this makes sense at all...
>>>
>>> Thanks in advance for the guidance,
>>> Ramon
>> --
>> Gromacs Users mailing list
>>
>> * Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>> posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>> send a mail to gmx-users-request at gromacs.org.