[gmx-users] creating a larger lipid membrane

[Mark Abraham]

Amit Choubey wrote:
> Dear Gromacs Users,
> 
> I have been trying to generate a larger lipid system but am running into 
> troubles due to improper initial co-ordinates set up. Following are the 
> steps i am trying to follow.
> 
> 1. I start with a 128 dppc membrane and use genconf to create a larger 
> system. I got the dppc128.pdb from Dr. Peter Tieleman's website.

What command line did you use?

> 2. I changed my topology file accordingly (change the number of DPPC and 
> Water molecules)
> 
> 3. I move on to energy minimization. During the mdrun i see the 
> following message :-

Test your protocol on the original coordinate file. It needs to work for 
  your procedure to be sound.

> Warning: 1-4 interaction between 1854 and 1857 at distance 6.249 which 
> is larger than the 1-4 table size 2.800 nm
> These are ignored for the rest of the simulation
> This usually means your system is exploding,
> if not, you should increase table-extension in your mdp file
> or with user tables increase the table size
> 
> I also notice that the max force does not converge to less than 2000 
> rather its of the order of 2e+06 .
> 
> 4. I tried to do energy minimization again on the above system but the 
> steepest descent could not change the potential energy substantially and 
> terminated after less that 20 steps. I again noticed the following warning 
> 
> Warning: 1-4 interaction between 1854 and 1857 at distance 5.269 which 
> is larger than the 1-4 table size 2.800 nm
> These are ignored for the rest of the simulation
> This usually means your system is exploding,
> if not, you should increase table-extension in your mdp file
> or with user tables increase the table size
> 
> 5. Now i changed my mdp files to do periodic_molecules = yes . Note pbc 
> = xyz has been used as default.
> 
> Same warnings were printed out during energy minimization.

You don't have periodic molecules. Changing mdp options haphazardly is a 
recipe for problems. Read 7.3 about them in the first instance, and then 
look up in the table of contents for more background.

> 6. I also tried to use the table-extension option and increased its 
> value to 10 nm but later the LINCS printed out too many warnings and MD 
> could not be done. I know that its not at all a good idea to increase 
> the 1-4 cut off.

Yup, bad idea. Fix the underlying problem, not the symptoms.

Mark