NAMD/PLUMED metadynamics: coordination number
Vaida
I have been working on ion in water system using metadynamics and use
2 collective variables - coordination number of ion with water
molecules in the first coordination sphere and coordination number of
ion with the rest of water molecules.
I have tried both - GROMACS/PLUMED and NAMD/PLUMED. While everything
worked fine with GROMACS, in NAMD the changes in coordination numbers
in the COLVAR file doesn't fit with the changes which I can observe in
the trajectory using vmd. While COLVAR file says that coordination
number changes by one, vmd shows only slightly increased/decreased
distance, but not the change in coordination number. While choosing
the parameters r0 and d0, I followed the advice specified in plumed
manual:
"In general a good first guess for these parameters can be achieved by
looking at
the pair distribution function and setting d0 equal to the position of
the first
peak in the pair distribution function, r0 as the full width at half
maximum
of the peak"
This is my plumed.dat file:
# general options
HILLS HEIGHT 0.01 W_STRIDE 2000
# print each 50 time units and add a time offset of 20.0 to COLVAR
PRINT W_STRIDE 50
# definition of CVs
COORD LIST <g1> <solvent1> NN 6 MM 12 D_0 2.20 R_0 0.2 SIGMA 0.01
g1->
1
g1<-
solvent1->
4 7 19 28 94 151
solvent1<-
COORD LIST <g2> <solvent2> NN 6 MM 12 D_0 2.20 R_0 0.2 SIGMA 0.01
g2->
1
g2<-
solvent2->
10 13 16 22 25 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82
85 88 91 97 100 103 106 109 112 115 118 121 124 127 130 133 136 139
142 145 148 154 157 160 163 166 169 172 175 178 181 184 187 190 193
solvent2<-
ENDMETA
I would really appreciate if you could help me with this issue..
Best wishes, Vaida
Fabio Pietrucci
could it be a problem of the units of distance which are different
in gromacs and namd (e.g. nm vs angstrom)? The parameters r0 and d0
have the units of the MD program you use.
Fabio
> --
> You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> To post to this group, send email to plumed...@googlegroups.com.
> To unsubscribe from this group, send email to plumed-users...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/plumed-users?hl=en.
>
>
--
____________________
CECAM Lausanne
homepage: http://sites.google.com/site/fabiopietrucci
Vaida
Thank you very much for your reply. I use nm in gromacs and angstroms
in namd. So this is not the cause of the problem..
Best wishes, Vaida
Giovanni Bussi
First a comment on the choice of CVs: clearly the indexes of the water
molecules in the first solvation shell are going to change during the
simulation. Thus, O am not sure if you can get what you want with this
CVs.
Concerning your problem, if I understand correctly you only have
problems with NAMD and not with GROMACS. This could be due to a bug in
PLUMED or simply to the random differences between the two
trajectories.
You should check the first CV, the second CV, and their sum (total
coordination). A change +1 in the first and -1 in the second simply
indicates an exchanged water molecule, an event which could happen at
different moments in the trajectory with two different codes.
Concerning units, also be careful with hills HEIGHT (that also goes in
the internal units of the code).
Giovanni
Vaida
Thank you very much for your reply. Actually, I want to observe water
exchange.
While in COLVAR file I can see the exchange - change in the
coordination number, I can't see it
while looking at the trajectory (in vmd).
The reason why I use two different programs: I would like to simulate
ion in 64 water molecules. However, in GROMACS it is quite difficult
to simulate such small systems, while it is easier in NAMD. Therefore,
I simulated ion in 64 waters with NAMD and ion in 891 waters with
gromacs.
Maybe the discrepancy arises due to the size of the system. I am
planning to try Zn + ~891 waters in NAMD and see if it changes
anything.
About HEIGHT. I checked that: in GROMACS we use kJ/mol and in NAMD -
kcal/mol. I took that into account.
Best wishes, Vaida
On Sep 16, 2:13 pm, Giovanni Bussi <giovanni.bu...@gmail.com> wrote:
> Hi Vaida,
>
> First a comment on the choice of CVs: clearly the indexes of the water
> molecules in the first solvation shell are going to change during the
> simulation. Thus, O am not sure if you can get what you want with this
> CVs.
>
> Concerning your problem, if I understand correctly you only have
> problems with NAMD and not with GROMACS. This could be due to a bug in
> PLUMED or simply to the random differences between the two
> trajectories.
>
> You should check the first CV, the second CV, and their sum (total
> coordination). A change +1 in the first and -1 in the second simply
> indicates an exchanged water molecule, an event which could happen at
> different moments in the trajectory with two different codes.
>
> Concerning units, also be careful with hills HEIGHT (that also goes in
> the internal units of the code).
>
> Giovanni
>