[gmx-users] Re: Problems with TIP4P

[Justin MacCallum]

Hi David,

thanks, that works now. I think the problem was that the special
water-optimized inner-loops were being used and the electrostatic
interactions weren't being calculated. Things seem to be fine if I take
my topology and flip the OW and MW around as you have.

Thanks,
Justin

> I've attached mine which works fine. (Just submitted some 50 ns
> simulations). This version is different in the sense that the order of
> atoms is flipped (Mw-Hw1-Hw2-Ow). It gives considerably better
> performance (by optimizing for quirks in gromacs innerloops).
> A short optimzation/equilibration of your starting coords should do >
for
> flipping the two atoms, as their original positions are quite close.

> Don't expect better results than gromos though...