How to restart a SMD simulation

LAMMPS LAMMPS Mailing List Mirror
1

Dear lammps-user,

I am trying to restart a SMD simulation which I perform a pull-out test of a protein.
In the initial SMD simulation, the tether point of the spring is far away from the SMD atoms which is named pulled.
The following is the command in the initial SMD simulation:

fix pull1 pulled smd cvel 10.0 -0.00001 tether 10000 NULL NULL 0.0

The output of the SMD vector is
Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6] pull1[7]
…
1100000 299.85343 -576509.82 6.966186 0 0 6.966186 9492.9317 9493.6591 64.079979

I restart the simulation at the 1100000th step.
In order to get the same force at the 1100000th step, I use the command shown below in the restart input file,

fix pull1 pulled smd cvel 10.0 -0.00001 tether 10000 NULL NULL 0.7274
(where R0=0.7274 = 9493.6591-9492.9317)

However, the force is not equal to 6.966186 and becomes negative.
In the output file:
Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6] pull1[7]
1100000 299.85343 -576509.82 -7.274 0 0 -7.274 9493.6284 9493.6591 64.079906

I think that problem is the value of R0 is not correct.

And also in LAMMPS manual, it mentions that LAMMPS store the information of SMD in restart file, and to re-enable it, the fix command in the new input script must use the same fix-ID and group-ID as was used in the input script that wrote the restart file.
But I have no idea how re-enable a SMD simulation in restart file.

Does anyone have some experience to restart a SMD simulation?
Or any idea why the force is not the same and even becomes negative?

Thank you.

Best,
Nina

2

Dear lammps-user,

I am trying to restart a SMD simulation which I perform a pull-out test of a
protein.
In the initial SMD simulation, the tether point of the spring is far away
from the SMD atoms which is named pulled.
The following is the command in the initial SMD simulation:

fix pull1 pulled smd cvel 10.0 -0.00001 tether 10000 NULL NULL 0.0

The output of the SMD vector is
Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6]
pull1[7]
...
1100000 299.85343 -576509.82 6.966186 0 0
6.966186 9492.9317 9493.6591 64.079979

I restart the simulation at the 1100000th step.
In order to get the same force at the 1100000th step, I use the command
shown below in the restart input file,

fix pull1 pulled smd cvel 10.0 -0.00001 tether 10000 NULL NULL 0.7274
(where R0=0.7274 = 9493.6591-9492.9317)

However, the force is not equal to 6.966186 and becomes negative.
In the output file:
Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6]
pull1[7]
1100000 299.85343 -576509.82 -7.274 0
0 -7.274 9493.6284 9493.6591 64.079906

I think that problem is the value of R0 is not correct.

no. your understanding or R0 is incorrect.

And also in LAMMPS manual, it mentions that LAMMPS store the information of
SMD in restart file, and to re-enable it, the fix command in the new input
script must use the same fix-ID and group-ID as was used in the input script
that wrote the restart file.
But I have no idea how re-enable a SMD simulation in restart file.

just use the exact same command. please read the fix smd documentation
about what R0 means. you almost certainly want it to be 0.0.

axel.

3

Axel can address the SMD-specific questions.

And also in LAMMPS manual, it mentions that LAMMPS store the information of
SMD in restart file, and to re-enable it, the fix command in the new input
script must use the same fix-ID and group-ID as was used in the input script
that wrote the restart file.
But I have no idea how re-enable a SMD simulation in restart file.

This simply means that you have to use a fix smd command
in the new simulation that is consistent with the fix
smd command you used in the 1st simulation.

Steve

4

Hi Axel

Thanks for your reply.
I used the exact same command as the initial one to restart the SMD simulation.

fix pull1 pulled smd cvel 10.0 -0.00001 tether 10000 NULL NULL 0.0

But in the output file, it shows:

Resetting global state of Fix pull1 Style smd from restart file info

it seems the SMD simulation is reset but not restarted.

Nina

2012/3/26 Axel Kohlmeyer <[email protected]>

5

Hi Axel

Thanks for your reply.
I used the exact same command as the initial one to restart the SMD
simulation.

fix pull1 pulled smd cvel 10.0 -0.00001 tether 10000 NULL NULL 0.0

But in the output file, it shows:

Resetting global state of Fix pull1 Style smd from restart file info

it seems the SMD simulation is reset but not restarted.

no. those are just words. look at the results.

axel.

6

Hi Axel,

The results also show the SMD is resetting not restarting. See below

Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6] pull1[7]
1100000 299.85343 -576509.82 0 0 0 0 9493.6284 9493.6591 64.079979
1100500 299.37591 -575789.56 3.1966622 0 0 3.1966622 9493.6234 9493.9728 64.089184

the force at the 1100000th step is reset to zero.

Nina

7

Hi Axel,

Sorry in my previous email I had not finished - this is a continued,

Although the force at the 1100000th step is reset to zero, I notice the value of PMF is the same as the last result in the initial SMD simulation.

I paste the result again please see below:

//======= The output of the SMD vector in the INITIAL simulation====
Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6] pull1[7]
…
1100000 299.85343 -576509.82 6.966186 0 0 6.966186 9492.9317 9493.6591 64.079979

//======= The output of the SMD vector in the RESTART simulation=====
Step Temp TotEng pull1[1] pull1[2] pull1[3] pull1[4] pull1[5] pull1[6] pull1[7]
1100000 299.85343 -576509.82 0 0 0 0 9493.6284 9493.6591 64.079979
1100500 299.37591 -575789.56 3.1966622 0 0 3.1966622 9493.6234 9493.9728 64.089184

Thank you very much.

Nina