[gmx-users] Do hydrogen bonds innately exist using mdrun?

[Justin Lemkul]

On 2/8/19 5:20 PM, James wrote:
> Hi,
>
> I know that gromacs has an hbond command which will analyze hbonds after a
> run. But, during a run, without doing something like specifying partial
> charges, do hbonds exist?
>
> It seems to me that they do not, because if I put two water molecules next
> to each other, they do not assume the configuration they would if hbonds
> were present, nor do the energies come out as expected (the affinity of the
> two water molecules is 10X too low, which makes me think that gromacs is
> using VDW, but ignoring hbonds).
>
> I'd love to be wrong and find out that hbonds "just work"! But, if they do
> not "just work," how do you specify them? I could see mimicking them by
> tuning partial charges on the donor and acceptor atoms to give the correct
> affinity, but it seems like that still would not be accurate because hbonds
> are angle dependent, whereas long range non-bonded forces are not.

There is no explicit energy function associated with hydrogen bonds. 
Many force fields used to include such a term, but it was found to be 
unnecessary with better parametrization. Hydrogen bonds arise due to 
favorable electrostatic interactions and are entirely a function of the 
water topology. Don't mess with topologies, because as you "fix" one 
thing, you break a whole host of others. There is no water model that 
achieves correct agreement with all structural, dynamic, and 
thermodynamic properties of water. Most do a good job of hydrogen 
bonding geometries, RDF, etc.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Office: 301 Fralin Hall
Lab: 303 Engel Hall

Virginia Tech Department of Biochemistry
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

==================================================