weight factor in force field file

The functionality of OpenMM will (eventually) include everything that one would need to run modern molecular simulation.
POST REPLY
User avatar
Ommair Ishaque
Posts: 11
Joined: Sun Feb 18, 2018 10:22 am

weight factor in force field file

Post by Ommair Ishaque » Tue Feb 20, 2018 9:14 am

Hi,
I have confusion in defining weight in force field file, what exactly weight is? I did read the http://docs.openmm.org/7.0.0/api-python ... eSite.html but fail to understand fully. How do we calculate these weight factors for some simple molecule like water tip5p? I am asking this because I will use this procedure for my system.
example given below:
<Atom name="O" type="tip5p-O"/>
<Atom name="H1" type="tip5p-H"/>
<Atom name="H2" type="tip5p-H"/>
<Atom name="M1" type="tip5p-M"/>
<Atom name="M2" type="tip5p-M"/>
<VirtualSite type="outOfPlane" siteName="M1" atomName1="O" atomName2="H1" atomName3="H2" weight12="-0.34490826" weight13="-0.34490826" weightCross="-6.4437903"/>
<VirtualSite type="outOfPlane" siteName="M2" atomName1="O" atomName2="H1" atomName3="H2" weight12="-0.34490826" weight13="-0.34490826" weightCross="6.4437903"/>

User avatar
Peter Eastman
Posts: 2593
Joined: Thu Aug 09, 2007 1:25 pm

Re: weight factor in force field file

Post by Peter Eastman » Tue Feb 20, 2018 10:51 am

The equations are given in a slightly more readable form at http://docs.openmm.org/latest/userguide ... tual-sites. But there's no single answer for how you choose them. The virtual site class defines how the site position is chosen based on the particle positions and the weights you provide. It doesn't specify how you should choose those weights.

In the case of an outOfPlane virtual site, weight12 and weight13 set the position within the plane of the three atoms. They give the fraction of the distance along the 1-2 and 1-3 bonds. And then weightCross determines the distance from the plane of the three atoms, as described by the formula.

User avatar
Ommair Ishaque
Posts: 11
Joined: Sun Feb 18, 2018 10:22 am

Re: weight factor in force field file

Post by Ommair Ishaque » Tue Feb 20, 2018 8:03 pm

Hi Peter,

Is it compulsory to define weight factor in force field field? I have water with 5 virtual sites with known position and angles of all sites(total 8 sites)?
As I mentioned in my previous post I have the functional form of the 3body potential energy which depends on the 5 parameters , and these parameters are different for different atoms interactions. I want to implement it using custom-non-bonded force option. Do you have any example for this? It this possible to implement?

User avatar
Ommair Ishaque
Posts: 11
Joined: Sun Feb 18, 2018 10:22 am

Re: weight factor in force field file

Post by Ommair Ishaque » Tue Feb 20, 2018 8:10 pm

I am attaching my potential energy form.

User avatar
Peter Eastman
Posts: 2593
Joined: Thu Aug 09, 2007 1:25 pm

Re: weight factor in force field file

Post by Peter Eastman » Wed Feb 21, 2018 11:16 am

The weights are part of the definition of the virtual site positions. They have nothing to do with the functional form of your interaction. If you have five virtual sites whose positions are computed based on the positions of three atoms, you must know how they're computed? You need to tell OpenMM that. <VirtualSite> tags are how you tell it that. It offers several different rules for computing virtual site positions, as described in the manual. You need to determine whether any of those rules can reproduce the computation you want it to do.

Once the virtual site positions are computed, you then need to compute the energy as a function of those positions. That's what the other tags in the file are for, for example <NonbondedForce> or <CustomNonbondedForce>.

POST REPLY