[gmx-users] Erik Lindahl,md_hole process!!

[Erik Lindahl]

On Thursday, Feb 20, 2003, at 19:24 US/Pacific, nanyu101 wrote:

> Dear Dr.Erik Lindahl,
>   Thanks for your prompt reply.I just want to know the process of 
> making hole.
> First,I should minimize my bilayer system such as dppc128.pdb or 
> popc128a.pdb downloaded from Dr.Tieleman's web.
> Second, I know my cylinder's radius.I run make_hole.pl to create a 
> hole as big as my cylinder.
> Third.I create my run.mdp and hp.mdp just as md_hole tutorial told.And 
> then run grompp -f run.mdp -o insert.tpr -c bilayer.pdb -r bilayer.pdb 
> -p bilayer.top
> mdrun -v -hole -holep hp.mdp -deffnm insert....
> If I manage the processes above,I can put my cylinder into my hole.
> cat protein.pdb bilayer.pdb > ion_bilayer.pdb.
>

Hi,

Sorry if it sounded like critizism, that wasn't my intention; it was 
just a word of warning that the programs will catch SOME of the obvious 
errors, but you have to be very careful and check the system after each 
step - it is quite easy to get other problems (like distorted protein 
structures) when inserting things into membranes, and the programs 
cannot detect many of them automatically.

Again, your last error message is indicative of atoms colliding or 
overlapping; the huge forces will lead to speeds above the velocity of 
light, and when you try to integrate this you will end up with NaN (not 
a number) values.  Check your coordinates, try the -debug flag and look 
into the log file to find the atoms with problems.

Cheers,

Erik