[gmx-users] Issues using Implicit Solvent with Charmm 27

[Juan José Galano Frutos]

Hi there,

I am trying to simulate a protein using a GBSA (implicit solvent) approach,
and first I've been googling a bit in order to set the parameters of the
simulation (mdp file) as better as possible.

I am using the following mdp options for the minimization step (the
parameters of the Implicit Solvent section are also set in the subsequent
NVT and NPT steps):
--------------------------------------------------------------------------------
title           = Energy Minimization   ; Title of run

; The following line tell the program the standard locations where to find
certain files
cpp             = /lib/cpp      ; Preprocessor

; Define can be used to control processes
define          =
cutoff_scheme   = group

; Parameters describing what to do, when to stop and what to save
integrator      = steep
emtol           = 1.0
emstep          = 0.01
dt              = 0.001
nsteps          = 40000
nstenergy       = 500
energygrps      = System

; Parameters describing how to find the neighbors of each atom and how to
calculate the interactions
ns_type         = simple
coulombtype     = cut-off
rcoulomb        = 0
vdwtype         = cut-off
rvdw            = 0
constraints     = none
pbc             = no
ld-seed         = 10000
nstlist         = 0
rlist           = 0
comm-mode       = angular
comm-grps       = Protein
optimize_fft    = yes

; Implicit solvent

implicit_solvent     = GBSA
gb_algorithm         = OBC
gb_obc_alpha         = 1
gb_obc_beta          = 0.8
gb_obc_gamma         = 4.85
gb_dielectric_offset = 0.009
nstgbradii           = 1
rgbradii             = 0
gb_epsilon_solvent   = 80
gb_saltconc          = 0
sa_algorithm         = ace-approximation
----------------------------------------------------------------------

The point is that after this minimization step I am obtaining a protein
where the alpha helixes seem to be something relaxed or extended in an
unusual way at least compared to when explicit solvent is used. This fact,
i.e. such a local structural weakening, lead then to a fast protein
unfolding during the subsequent NVT and NPT steps something that I am sure
is not normal that soon.
So, my question is whether there is someone over there has performed
simulations combining CHARMM27 force field and a GBSA approach and has
happened the same? Is there any incompatibility between them using GROMACS?
I could find some papers in which this combination is used but using NAMD.
Any help or advise would be appreciated so much.

Thank you!!


Juan José Galano Frutos

Department of Biochemistry and
Molecular and Cellular Biology,
Faculty of Sciences,
University of Zaragoza
Pedro Cerbuna # 12, 50009
Zaragoza (Spain)
+34 976 76 28 06

Institute for Biocomputation and
Physics of Complex Systems (BIFI)
Mariano Esquillor, Edificio I + D - 50018
Zaragoza (Spain)