Re: amber-developers: rna pdb atom map questions

From: David A. Case <>
Date: Fri, 23 Mar 2007 09:05:42 -0700

On Thu, Mar 22, 2007, Scott Brozell wrote:

> ...
> Created a new atom named: H2* within residue: .R<RC3 2>

Well, the current PdbAtomMap for RNA has this line:

  { "H2*1" "H2'1" }

But I see that some (not all!) pdb RNA files use H2* instead of H2*1. For
example, 28sp.pdb uses the H2*1 name, but 1jzc.pdb uses just H2*.

The new "remediated" pdb (see uses H2' for this hydrogen atom,
whereas Amber uses H2'1. At some point, we will have to switch over, probably
at the switch to Amber 10.

> Why is it that none of these atom maps fixes the naming problem
> addPdbAtomMap { { "H2*" "1H2*" } }
> addPdbAtomMap { { "H2*" "H2*1" } }
> addPdbAtomMap { { "H2*" "1H2'" } }
> addPdbAtomMap { { "H2*" "1H2" } }
> addPdbAtomMap { { "H2*" "H2'" } }
> but that this one does
> addPdbAtomMap { { "H2*" "H2'1" } }

Because the Amber atom name (currently) is "H2'1", and only the final example
converts the input to that name. The output (second item) of an addPdbAtomMap
entry must correspond to what is in the amber library files.

> Note that leap's savepdb uses the atom name 1H2*.

This is not correct. Savepdb converts H2'1 to H2*1, and then encodes the name
so that the atomic symbol is in the second column. This "pdb-file encoding"
of names is going away with the new pdb, and should reduce lots of confusion.
Also going away is the use of "*" rather than "'" in nucleic acid sugars.

> Also what are the perils of always using
> addPdbAtomMap { { "H2*" "H2'1" } }
> when reading rna pdb's ?

It looks like this should be done (i.e. added to the leaprc.rna.* files). I'm
not sure right now why no-one has tripped over this before. Probably because
only a few RNA pdb files come with hydrogens attached, and many/most of those
use H2*1 instead of H2*.

Thanks for the report....dac
Received on Sun Mar 25 2007 - 06:07:45 PDT
Custom Search