[gmx-users] 1-4 Interactions

[Mark Abraham]

toma0052 wrote:
> Hi,
>      Thanks for the prompt responses.  I checked the distance between atoms
> 908 and 913 in VMD and got 0.396nm, which seems fine.  I thought then, that
> my problem may be with my box size when I used editconf.  When I changed
> the box dimension to 2.0nm from the molecule periphery, the warning between
> atoms 908 and 913 disappeared.  However, I then recieved a similar error 
> 
> warning: interaction between 911 and 932 at distance 6.244 which is larger
> than the 1-4 table size 1.000 nm
> 
> I checked atoms 911 and 932 using VMD, and they don't appear to be close at
> all, and their distance is in fact 6.244nm.  Does this mean that there is a
> problem with my topology file? 

Yep - as I guessed earlier, I'd still guess your topology doesn't match 
your starting structure.

> My topology file is the same as that given
> on Tieleman's site, except that I removed #include "lipid.itp".  This was
> giving me an error, and I did download the ffgmx_lipids.tar.gz file from
> the Gromacs site, which seemed to solve the problem. 

These things are very brittle - removing an #include is one of the 
simplest recipes for a problem I can think of. Fix this problem, and if 
you still have problems, then run cpp on the .top file to pull in all 
the #inclusions and go through the result with a fine-tooth comb and see 
that it all makes sense with respect to your structure.

> So, I am not sure why
> this error is occuring.  The problem could be with my em.mdp file. 

Nah, looks fine, except that pbc=full isn't defined in the manual. I do 
remember hearing about it once, but perhaps you should try pbc=xyz instead.

Mark