[gmx-users] Constraints and EM

[irene farabella]

Dear all,

I have some doubt regarding energy minimization and the use of
constraint algorithms (such as LINCS) that may come from my
inexperience in GROMACS and molecular dynamics in general.

In a Berk Hess talk given at the GROMACS workshop 2007 it is stated:

(https://extras.csc.fi/chem/courses/gmx2007/Berk_talks/forcef.pdf)

"In GROMACS bonds are usually replaced by constraints, i.e. the
distances between bonded atoms are fixed to their equilibrium value.
Constraints are actually a more faithful representation of reality
since the dynamics of bonds is of quantum nature..”

In an article about the use of SHAKE (another algorithm for
constraints) I read that:
"Iinitiating a constrained simulation from an unconstrained minimum
usually results in spurious temperature fluctuations that slow down
the equilibration of the molecule and/or introduce artifacts" (Duan,
Y., Kumar, S., Rosenberg, J.M. and Kollman, P.A. "Gradient SHAKE: An
improved method for constrained energy minimization in macromolecular
simulations." Journal of Computational Chemistry (2004), 16(11):
1351-1356.)

That suggests to me that the use of constraints during EM is somehow suggested.
However why use constraints (that maintain fixed lengths of the
selected bonds) during this step of the simulation if what you want to
achieve with EM is the local potential energy minimum and it is
required to have a high degree of freedom in the atoms' movement?

Is my interpretation of the use of constraints wrong?

Sorry for the long email, any suggestion from an expert point of view
is welcome.

Thanks

-- 
Irene Farabella
Wellcome Trust PhD student

Institute of Structural and Molecular Biology
Department of of Biological Sciences
Birkbeck
University of London
Malet Street
London WC1E 7HX
Telephone +44 (0)20 7631 6815