Re: Some questions about the orthogonality of colvars in ABF.

From: Jérôme Hénin (jerome.henin_at_ibpc.fr)
Date: Wed Mar 15 2017 - 07:26:23 CDT

Hi yjcoshc,

> Does that means the oneSiteTotalForce option only affect the measurement
> of the forces and doesn't have effect when applying the forces to atoms?
>

That is correct.

> If two colvars have overlapping atoms and they apply the bias forces to
> their common atoms, are these two colvars still mutual orthogonality?

Yes, as long as they don't measure forces on those common atoms. The exact
requirement is that the gradient of one variable is orthogonal to the
inverse gradient of the other [1], or equivalently, that biasing forces on
one variable are orthogonal to the force measurement on the other.

Best,
Jerome

[1] Exploring Multidimensional Free Energy Landscapes Using Time-Dependent
Biases on Collective Variables, J. Chem. Theory Comput., 2010, 6 (1), pp
35–47
http://pubs.acs.org/doi/abs/10.1021/ct9004432

On 15 March 2017 at 13:11, yjcoshc <yjcoshc_at_gmail.com> wrote:

> Dear All,
>
> I am confused about the orthogonality of colvars and the oneSiteTotalForce
> option in the user guide. The user guide says:
>
> atoms involved in the force measurement on i do not participate in the
> definition of j . This can be obtained using the option oneSiteTotalForce
> of the distance, angle, and dihedral components (example: Ramachandran
> angles , ).
>
> Does that means the oneSiteTotalForce option only affect the measurement
> of the forces and doesn't have effect when applying the forces to atoms?
> If two colvars have overlapping atoms and they apply the bias forces to
> their common atoms, are these two colvars still mutual orthogonality?
>
> Thanks,
> yjcoshc
>
>

This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:09 CST