[gmx-users] simulating Cellulose using GAFF force field

[Fahimeh]

Hi gromacs users,

I am trying to melt a crystalline structure of cellulose. This structure contains 8mers and monomers of cellulose molecules. To start the simulation, I computed the resp charges of 8mers and monomers of Cellulose and I am using GAFF force field for simulating this structure. To test wether the force field is correct … I performed a simulation using only one 8mer molecule and I it doing well. I can bring the temperature up to 300k and at that point the cellulose molecule is folded and seems nothing is wrong … however when I start to run the simulation for the crystalline Cellulose the problem starts.

First I minimized the system using steepest decent and then I perform NVT at low T(10k) which is followed by NPT simulation to warm the system from 10k to 298k. I increase the temperature every 1000ps by 10 degree which is very slow based on my experience … But the system explode every time the temperature reaches above 150 or 200k. Using dt=0.001fs and different types of barostate doesn’t help as well. 

I thought the problem might be the starting configuration. Therefore I used the relaxed 8mer that I initially simulated … and I used packmol program to put almost 50 of 8mers in a box based on the experimental density that I had. Packmol successfully pack the relaxed 8mers in the box and then I repeat all the steps that I explained above… but exactly the same problem happens. 

At this point I am not sure where the problem can be:
1- If the problem is force filed … .then how comes that I can run the simulation of one 8mer molecule without any problem
2- If the problem is starting configuration, then how comes that even using other techniques such as packing the relaxed molecules in a box … does not help

It would be great to hear your ideas

Thanks
Fahimeh