[gmx-users] Re: genbox problem with micelle

[chris.neale at utoronto.ca]

> A third alternative would be to increase the vdw radii, by copying
> vdwradii.dat to your working directory and increase the water radius.

Actually I find that increasing the radius of water can cause the  
overall water to be less dense even far away from the protein/micelle.  
I do suggest copying vdwradii.dat to your working directory but then  
add lines to ensure that the default water radius does not get too  
large, e.g. for spc:

??? OW 0.105
??? HW1 0.04
??? HW2 0.04

And then use the -vdwd flag with a value larger than 0.105. This way  
your water appears normally to the new waters but the micelle atoms  
all seem larger.

Keep repeating with -vdwd values that are slowly getting larger until  
you find the minimal value of -vdwd for which no waters are placed  
inside the micelle.

Also, how long did you simulate? The water should actually get  
squeezed out within 5 ns, and for the size of micelle that has been  
simulated in the past you should be able to get 0.5ns-1ns/day.

Chris.