[gmx-users] Hydroxyapatite MD odd behavior

[Ramon Garduno]

Dear gmx friends:

We are modeling a solvated hydroxyapatite(HAP)-protein complex. We have
minimized and equilibrated this system to 300K under NVT.

The odd behavior is observed when we start an NPT simulation. In the very
first ps we observed that the oxygens bound (initially) to phosphorous in
phosphate tetrahedon assume a geometry in which all four oxygens are grouped
together like this P=O, however we never receive an error message.

The bond lenghts have been constrained using the SHAKE algorithm.

HAP is in the form of a crystal slab. The HAP parameters we have used were
taken from Dinesh-Katti's paper [Polymer 48 (2007) 664-674], in turn, Katti
took the HAP parameters from Hauptmann [Phys Chem Chem Phys  (2003) 5, 635-639]. 

Katti's parameters were parameterized to CVFF. We have made ​​the
corresponding unit convertions to fit those requiered in GROMACS v4.5. In
short, we have used NON_BONDED (Homonuclear and Heteronuclear van der Waals
terms), BONDED parameters (ffnonbonded.itp and HAP.itp), and the same PARTIAL
ATOMIC CHARGES as those reported by Katti's group.

In order to avoid the odd behaviour observed in the tetrahedral phosphate, we
have used strong position restraints but the odd behaviour is still present.
We also used freeze groups but with this command it is not possible to
equilibrate the system under NPT. 

Do you have some suggestions to solve this problem?

I Look forward to receiving your comments.

Best,

Ramon Garduno
Research Professor in Biophysics
Universidad Nacional Autonoma de Mexico

--
Open WebMail Project (http://openwebmail.org)