[gmx-users] Hydrogen database
Mark Abraham
Mark.Abraham at anu.edu.au
Fri Nov 18 15:46:04 CET 2011
On 18/11/2011 3:11 AM, Ehud Schreiber wrote:
>
> Hi,
>
> I have recently studied the hydrogen database format of .hdb files
> (page 118, section 5.6.4 in the manual version 4.5.4). I would like to
> make a few remarks that, if correct, may need addressing.
>
> 1)Method 3 of adding the hydrogens, that of two planar hydrogens,
> gives -NH2 as the example. I think this is misleading, as although
> this is true for an amide group --C(=O)NH2 such as in an asparagine
> and glutamine side chains, the nitrogen is tetrahedral in the R-NH2
> case or in the amino acid N-terminus.
>
True, though attempting to model pyramidal inversion with the usual MM
tool kit sounds like a recipe for trouble anyway.
> A better example for two planar hydrogens would be =CH2 such as in
> ethylene or vinyls.
>
True, I'll add that to the manual ahead of amide -NH2.
> 2)The provided methods for adding hydrogens are not covering the whole
> set of possibilities. In particular, it seems to me that three methods
> are lacking, although admittedly they are less common:
>
> a.One tetrahedral hydrogen connected to atom i which is in turn
> connected to two atoms j,k such that n is on the plane bisecting angle
> j-i-k; n-i-j = n-i-k = 109.47 degrees; and dihedral n-i-j-l > 90
> degrees. Example: secondary amines R2NH. This case can be mimicked by
> method 2 with i,j,l atoms so is perhaps superfluous.
>
> b.One planar hydrogen connected to atom i which is connected to only
> one other atom j such that n-i-j = 120 degrees and n-i-j-k is trans.
> Example: R2C=NH.
>
Yeah I'd use method 2 for both of these and rely on EM to fix it, but
some hints are in order.
> c.One linear hydrogen such that n-i-j is a straight line. Example: #CH
> where # is a triple bond.
>
That's a potentially useful feature, but method 2 would still be
acceptable with due care.
> 3)I haven't checked this, but can the k atom be a hydrogen added in an
> earlier line of the same .hdb file?
>
I would think so.
Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20111119/0e9502fa/attachment.html>
More information about the gromacs.org_gmx-users
mailing list