Martini Force Field for carbohydrates

LAMMPS LAMMPS General Discussion
1

Hi all,
I am trying to simulate coarse grained cellulose chain using martini force field parameters.
I used the following parmeters from this Paper (https://pubs.acs.org/doi/10.1021/jp5105938)

pair_coeff 	1 1 1.33 4.7 
pair_coeff 	2 2 0.478 6.2
pair_coeff 	3 3 1.33 4.7 
pair_coeff 	4 4 1.33 4.7
pair_coeff 	5 5 0.478 6.2 
pair_coeff 	6 6 1.33 4.7
bond_coeff 	1 71.7 2.32
bond_coeff 	2 71.7 2.22
bond_coeff 	3 71.7 5.55 
bond_coeff 	4 71.7 2.20
bond_coeff 	5 71.7 2.30 

angle_coeff 	1 11.95 166
angle_coeff 	2 11.95 85
angle_coeff 	3 11.95 85
angle_coeff 	4 11.95 113
angle_coeff     5 11.95 80
angle_coeff     6 11.95 165

dihedral_coeff 1 2.39 1 0 
dihedral_coeff 2 2.39 1 0

The (.in) file is given below:

    units         real
     dimension     3
     boundary     p p p
     atom_style   full
    
#     neighbor       2.5 bin
#     neigh_modify   delay 5
    
pair_style	lj/cut 12.0
#pair_modify 	mix arithmetic
pair_modify    shift yes mix geometric
bond_style	harmonic
angle_style	harmonic
dihedral_style  harmonic

read_data	cell5.lmp
include         Parameter.lammps


# Assign original velocities to atoms
compute         new all temp
velocity        all create 298.1 487639 temp new

# Set up ensemble 
 fix             1 all nvt temp 298.1 298.1 100.0
timestep        5

# Output temperature, energy, pressure of the system 

     thermo_style    custom step temp etotal pe ke pxx pyy pzz
     thermo_modify   flush yes
     thermo          100
     dump            1 all xyz 100 dumpcg13.xyz
run             200000
    
     write_data     cg13.data

But, I am not getting the negative potential energy. Do you have any suugestions?

Thanks,
Pinky

2

Why should the potential energy be negative?

3

Hello Sir,
How should I validate that my CG model is correct? I was trying to compare the energy curve of AA and CG model.

Sincerely,
Pinky

4

How to validate your coarse grained structure is a question for the documentation of the CG force field that you are using and publications describing applications.

Absolute values of the (potential) energy are arbitrary in classical models, only energy differences matter, however due to the coarse graining procedure, especially if it is a more “phenomenological” approach like MARTINI, there is not that much of a direct equivalence of interactions.

Instead you need to look are more indirect properties and even then there are multiple problems like the uncertainty of the time scale of the coarse grained system (structural changes tend to happen much faster because there is less “jamming” than in an all-atom model due to the less rugged and smoother potential hypersurface).