Uppsala Software News @ 960521
                ******************************


* MANUALS
*********
The text of the "News from the Uppsala Software Factory" articles from
previous CCP4 Newsletters is now available as part of the complete set
of HTML manuals.  Get pub/gerard/html_manuals.dirtar.Z from the server.
The main page in this set (gerard_manuals.html) now also contains links
to currently available pre-prints.


* HETERO
********
A new version of the hetero-compound database is now available (as
OMAC/hetero.pdb).  It contains coordinates for 1367 uniquely named
hetero entities (but note that calcium ions go under 22 different
names in the PDB [and counting ...]).  See also CT2HET below for a
new way of finding if anything useful (for you) is in this database.


* SYMBOLS
*********
All programs which use the command-line-loop interface now have a
simple mechanism for using symbols.  This is probably mostly useful
for people who want to write macros.  The programs are: MAMA (3.8),
MAPMAN (4.6), DATAMAN (4.1), LSQMAN (4.6), MOLEMAN2 (1.0) and
ODBMAN (1.4).


* MOLEMAN2 Version 1.0
**********
By popular demand (and long overdue): the SPlit command (which splits
your molecule into separate PDB files and generates an X-PLOR
generate.inp file) will now do two more things for every C-terminal
protein residue:
- make sure that the O or OT1 atom is written as OT1
- check if there is an OT2, OTX or OXT atom and rename it to OT2; if
  there is no such atom, an OT2 atom is generated on the fly and
  appended to the PDB file (but not added to the structure in memory).
This means that the process of preparing PDB files for use with X-PLOR
should now be free of manual PDB-file editing ("peppar, peppar").
  The program now also checks for atoms which have X ~ Y ~ Z; often,
this indicates atoms with "unknown" coordinates (e.g., (1500,1500,1500)
or (0,0,0)), which may lead to all sorts of trouble later on (e.g.,
when expanding a map around your molecule).  This check is done when
a new molecule is read, and when the BOokkeep command is executed.
  New PDb CHemical+charge command.  In the new PDB format definition,
column 77-78 of each ATOM and HETATM record contains the symbol of the
chemical element type of the atom, and column 79-80 its charge (don't ask
me why the charge has to fit inside two columns ...).  This command tries
to deduce this information as best as it can, using information in the
atom name (4 characters) field.
  Added optional "use_masses" parameter to the XYz CEntre_origin command.
For this to work, the program must be able to deduce the chemical
element symbol for all selected atoms (see previous paragraph).


* CT2HET Version 0.1
********
This is a little jiffy program that may be of help when you are looking
for all PDB entries that contain a certain type of hetero compound, or
to see if there is any hetero compound in the PDB which is a
superstructure of your particular one (for instance, if you want to
generate O or X-PLOR dictionaries).
The program requires two input files:
- a description of a hetero compound in connection table format
  (either generated by something like ChemDraw, or created by
  editing a file);
- the collection of hetero compounds available from Uppsala
  (OMAC/hetero.pdb).
The nice thing is that you only have to specify atom types and bonds in
your connection table file, and that it will also find hits for compounds
with trivial names (such as "Hoechst 33258" or "See Remark 6") which Unix
tools such as grep and awk would not come up with.


* LSQMAN Version 4.6
********
New HIsto_disto command to print a histogram of the distances of the
atoms in one molecule to the nearest atom in another molecule after
applying the current RT-operator.  This has the advantage that there
doesn't have to be a 1:1 correspondence in the atoms compared (so you
can compare a correct and a backwards traced model, for instance).  It
has the same disadvantages as all other distance-based measures
(sensitive to operator, poor statistic when domain shifts occur).


* COMING SOON
*************
A paper describing ESSENS has been submitted.  Once accepted, the
program will be released.  It does a very good job at finding the
helices and strands (or any other rigid entities) in your experimental
MIR (or other) map.  If you're curious, read the manual at URL:
   http://onyx.bmc.uu.se/~gerard/manuals/essens_man.html


* NOT COMING SOON AT ALL
************************
The next "Uppsala Software News".  Over the next five months I will
be travelling most of the time.  Will read my mails, but won't be able
to do too much in this period.


--cd