CCP4 SFALL program

Dave Schuller schuller at indigo1.biomol.uci.edu
Mon Jan 23 18:19:41 EST 1995


gerard at rigel.bmc.uu.se (Gerard Kleijwegt) wrote:
<I think it's plain old PDB atom names, with one important
<exception: the atom names are four characters, of which
<the *first two* (right justified) are used for the symbol
<of the chemical.  Therefore, ' CA ' is an alpha-carbon
<atom, whereas 'CA  ' would be a calcium atom !  In the
<same vein, ' NE ' is an epsilon-nitrogen, but 'NE  ' would
<be a neon atom, etc ad infinitum.

The official "Protein Data Bank Atomic Coordinate and
Bibliographic Entry Format Description" is available
as postscript file pub/format.desc.ps via anonymous
ftp from pdb.pdb.bnl.gov . This is reccomended reading for
anyone who writes or uses crystallographic programs. The
guideline you mention in which the first 2 letters of the
atom name are the element name appears here (pp 26,28). So do
the PDB standard atom names for coenzyme A (p. 31) and
nicotinamide adenine dinucleotide (p.35), which do not follow
this rule.

also:
<As for SFALL, I haven't checked the source, but I assume
<it checks the first two characters of every atom name
<to deduce the chemical type.  If you have anything out
<of the ordinary, include a FORM(factor) line to tell
<the program for which chemicals it should look up
<form factors, e.g.:
<FORM C H N O S P CA
<if you would have a Ca and a PO4 in your structure.

you should have checked the source. sfall does indeed use
an array of 'standard' atom names rather than pulling out the
element name. It is sensitive to column placement. It does allow
any unusual atom name to be used if the atomic number is written
into columns 68:70 of the PDB record. I have a quickie program to
do this for the standard NAD atom names, which could be modified
for any special case. It can be obtained via anonymous ftp from
indigo4.biomol.uci.edu . Get pub/pdb/tar.Z .

David J. Schuller
University of California-Irvine




More information about the Xtal-log mailing list