Re: [AMBER] Diffusion in pmemd.cuda

From: David A Case via AMBER <>
Date: Fri, 9 Jun 2023 08:28:54 -0400

On Thu, Jun 08, 2023, David Poole via AMBER wrote:
>Interestingly, despite this being in line with literature procedures, the
>resulting diffusion constants calculated with the velocity autocorrelation
>function or diffusion analysis in cpptraj did not match literature nor was
>there an expected correlation between the molecule size and diffusion

Can you say a little more here? Were the (translational?) diffusion
constants too big or to small? What were the smallest and largest molecules
you used? Which water models were used? Can you provide an example of a
computed results compared to its target (experimental) value?

Generally, almost all common biomolecular water models have viscosities that
are too low. Do you see that TIP3P gives faster diffusion than something
like TIP4Pew or OPC? Have you computed the diffusion constant of water in
water? How does that compare to literature values (both experimental and
calculated)? I'm no expert here, but you might check out this paper, if
you've not already read it:

%A H. Zhang
%A C. Yin
%A Y. Jiang
%A D. van\0der\0Spoel
%T Force Field Benchmark of Amino Acids: I. Hydration and Diffusion in Different Water Models
%J J. Chem. Inf. Model.
%V 58
%P 1037-1052
%D 2018

Diffusion of (especially) small molecules is often dominated by hydrogen
bonding interactions between the solute and water. These are mostly
parameterized against free energies of solvation, but that might not be
enough to get good diffusion constants.

Finally, only NVE calculations should really be considered for careful
diffusion measurements. Some thermostats will have only small effects on
time-dependent properties, but you need to take care in how you choose the
thermostat parameters for NVT or NPT.

There are probably better papers than the one cited above, and I hope that
people on this list more direct experience will chime in. (I've only looked
at rotational diffusion of proteins of various sizes.)


AMBER mailing list
Received on Fri Jun 09 2023 - 05:30:02 PDT
Custom Search