van der Waals distances

From: David Mueller
Date: 20 March 2012 14:41

Dear Crystallographers:

I am mapping out van der Waals contacts between a small molecule and a protein.  What are the generally accepted limits for positive van der Waals interactions?  I am defining van der Waals interactions as the energy of interaction between permanent dipoles, plus the energy of interaction between permanent dipoles and dipoles induced by them, plus the energy of interaction between neutral atoms.  Is this complete? Accurate?


Is there an associated reference?

Cheers


David M. Mueller

----------
From: Rubén Sánchez Eugenia
Hi David Mueller,

In Physical Chemistry Van der Waals interacions are defined as all type of forces between molecules (or parts of them) excluding covalent bonds and electrostatic interactions. So, you are right that the most common forces included into Van Der Waals are:  interaction between permanent dipoles, interaction
between permanent dipoles and dipoles induced by them, and interaction between apolar groups due to their polarizability.

This may appear in the next reference (IUPAC Glossary of terms):
http://iupac.org/publications/pac/66/5/1077/

Regarding the subject that says "van der waals distance" this is the half of the distance of two non-covalent-bounded atoms. This specific distance I supposed it is due to the Van der Waals interactions, but maybe there are other causes.

Best regards.

--
---------------------------------------------------
Rubén Sánchez


----------
From: Ed Pozharski

On Wed, 2012-03-21 at 10:16 +0100, Rubén Sánchez Eugenia wrote:
> In Physical Chemistry Van der Waals interacions are defined as all
> type of forces between molecules (or parts of them) excluding covalent
> bonds and electrostatic interactions. So, you are right that the most
> common forces included into Van Der Waals are:  interaction between
> permanent dipoles, interaction
> between permanent dipoles and dipoles induced by them, and interaction
> between apolar groups due to their polarizability.

But the dipole-related interactions are electrostatic! The full quote
is


> The attractive or repulsive forces between molecular entities (or between groups within the
> same molecular entity) other than those due to bond formation or to the electrostatic inter-
> action of ions or of ionic groups with one another or with neutral molecules. The term includes:
> dipole-dipole, dipole-induced dipole and London (instantaneous induced dipole-induced dipole)
> forces.

It also says

> The term is sometimes used loosely for the totality of nonspecific attractive or repulsive
> intermolecular forces.

which is, indeed, a loose (and vague) definition (what is meant by nonspecific?).

The van der Waals forces arise, imho, from an attempt to break the
overall of electrostatic interactions into manageable components. This
neglects higher terms in multipole expansion.  Difficulty is that it
also includes contact repulsion.

To answer the original question, you may find the following threads useful

http://www.dl.ac.uk/list-archive-public/ccp4bb/msg19876.html
http://www.dl.ac.uk/list-archive-public/ccp4bb/msg19861.html

Ligplot is an excellent reference point, and it traces back to the
hbplus which you can use to output the list of non-bonded
interactions.  There is als the ligand explorer in the PDB which you
probably can use with your own model.

Normally, van der Waals contact distances are defined as the sum of the
van der Waals radii of the contacting atoms, so it's not a single
number.

Cheers,

Ed.

--
Oh, suddenly throwing a giraffe into a volcano to make water is crazy?
                                               Julian, King of Lemurs

----------
From: David Mueller
Aren't all forces between atoms and molecules electrostatic in origin?


Cheers

----------
From: <eugene.krissinel

plus exchange interaction :)

----------
From: Sheriff, Steven

David:

In my experience most people don't put a lot of thought into an upper limit for van der Waals (VDW) contacts and use a "hard", arbitrary upper value of 4 Angstrom.

In the mid-1980's, Wayne Hendrickson and I proposed that one should use:

·         Atom type dependent cutoffs
·         As an approximation to the asymptote to zero energy at infinite distance, an upper bound that was as much above the distance of minimum energy (Rmin) as Rmin was above the zero energy, Rzero, which is the point at which the repulsive forces exceed the attractive forces. Using a form of a 6-12 potential, we calculated that Rlimit = 1.11 x Rmin (and that Rzero = 0.89 x Rmin).

I have used that in all my succeeding work, e.g. on antibodies, although during a stint in David Davies' group at NIH following working with Wayne at the Naval Research Laboratory, I switched from using the very simple 4 atom types that PROTIN used to a somewhat more complex set of values that I was introduced to based on a paper by Gelin & Karplus (1979).

Here are some references:

S. Sheriff, W. A. Hendrickson & J. L. Smith (1987).  The Structure of Myohemerythrin in the Azidomet State at 1.7/1.3 Å Resolution.  J. Mol. Biol. 197, 273-296.

S. Sheriff (1993).  Some Methods for Examining the Interaction between Two Molecules.  Immunomethods 3, 191-196.

B. R. Gelin & M. Karplus (1979).  Side-Chain Torsional Potentials: Effect of Dipeptide, Protein, and Solvent Environment.  Biochemistry 18, 1256-1268.

And here are some reviews that used that used this calculation to compare structures:

D. R. Davies, E. A. Padlan & S. Sheriff (1990).  Antibody-Antigen Complexes.  Annu. Rev. Biochem. 59, 439-473.

S. Sheriff (1993). Antibody - Protein Complexes.  Immunomethods 3, 222-227.

Papers on antibody/antigen complexes that were edited by Ian Wilson in his capacity as an editor of J. Mol. Biol. were often requested to use the methodology as well.  Unfortunately [ ;) ], too many papers appeared outside of J. Mol. Biol. that were ignorant of the methodology put in place by me and used by Ian's group at Scripps as evidenced by (although these report on buried surface area calculations):

I.  A. Wilson & R. L. Stanfield (1993). Antibody-antigen interactions (1993). Curr. Opin. Struct. Biol. 3, 113-118.

I.  A. Wilson & R. L. Stanfield (1994). Antibody-antigen interactions: new structures and new conformational changes (1994). Curr. Opin. Struct. Biol. 4, 857-867.

for these methods to remain the standard in the field.

In the department of totally shameless self-promotion (as opposed to the above shameless self-promotion), I have published a recent paper, where I have used this methodology for the field of 10Fn3-based variants, when used as binding proteins in a manner analogous to antibodies, or more similarly, VHH domains (Camelid-like VH domains):

V. Ramamurthy, S. R. Krystek, Jr., A. Bush, A. Wei, S. Emanuel, R. DasGupta, A. Janjua, Z. Lin, L. Cheng, M. Murdock, D. Cohen, P. Morin, J. H. Davis, M. Dabritz, D. C. McLaughlin, K. A. Russo, G. Chao, M. C. Wright, V. A. Jenny, L. J. Engle, E. Furfine & S. Sheriff (2012).  Structures of Adnectin/Protein complexes reveal an expanded binding footprint.  Structure, 20, 259-269.

Steven

----------
From: Ed Pozharski
Technically, no.  You may be able to exclude nuclear forces, but gravity
certainly isn't included in Maxwell's equations.

The other forces can simply be neglected because their contribution is
negligible when molecular interactions are concerned.

On Wed, 2012-03-21 at 08:42 -0500, David Mueller wrote:
> Aren't all forces between atoms and molecules electrostatic in origin?

----------
From: Michael Schnieders
Hi David,

Perhaps the most accurate inter- and intramolecular forces available for
macromolecular X-ray crystallography refinement are currently based on the
polarizable multipole AMOEBA force field. AMOEBA includes explicit support
for the interactions you mentioned:

1.) repulsion-dispersion (Pauli repulsion and London dispersion forces)

2.) permanent multipole - permanent multipole (including permanent dipole
- permanent dipole)
3.) permanent multipole - induced dipole (including permanent dipole -
induced dipole)
4.) induced dipole - induced dipole (many-body polarization)

Secondary quantum mechanical effects are either implicitly accounted for
in the model when the repulsion-dispersion term is fit to experimental
thermodynamics (liquid densities and heats of vaporization) or are simply
neglected. Although AMOEBA is more expensive than traditional chemical
restraints used for refinement, we've shown it's fast enough for any data
set in the Protein Databank (i.e. even Ribosomes with 2 million atoms in
the unit cell).

The inter- and intramolecular distances achieved by AMOEBA-assisted X-ray
refinement have been shown to be very accurate. For example, in a recent
study of 17 macromolecular datasets we achieved an average MolProbity
score in the 95th percentile (see the reference below and citations within
for more details).

AMOEBA-assisted X-ray refinement is implemented in a free, open source
(GPL v.3) program called "Force Field X". The FFX engine could be included
in existing refinement programs if there is interest and we are open to
considering an alternative license as necessary.

http://ffx.kenai.com

Best Wishes,
Mike

Fenn, T. D. and M. J. Schnieders, Polarizable atomic multipole X-ray
refinement: Weighting schemes for macromolecular diffraction, Acta
Crystallographica Section D 2011 67, (11), 957-65.
http://dx.doi.org/10.1107/S0907444911039060