MM summary
The summary for the MM
Dear netters:
Thanks very much for the useful comments and suggestions of answering
my e mail questions:
> A friend of mine is going to start her master degree soon. Part of
> her work is to do some calculations on some crystal structure by MM2 or
> MM3(Molecular Mechanics). She will be apreciated to listen to some comments
> about MM2 and MM3 made by computational chemistist. Does anyone know well
> about the MM's?
>
My friend and I are appreciated. It could be very helpful for her
further reaserch.
I myself is majoring some theoretical work on the rovibrational states
of small molecules. I wish I could be able to get some suggestions and help
from this group. I will quote them in my thesis.
Thanks very much and best wishes.
Feng
--
| Feng WANG Department of Chemistry, The University of Newcastle |
| Callaghan, N.S.W.2308, Australia |
| Internet: chfw (+ at +) cc.newcastle.edu.au Tel: +61 49
215 475 |
|______________________________________________________________________________
1. Dr. Sauser wrote:
Wolfgang Sauer
Institut fuer Organische Chemie
Henkestr. 42, D-8520 Erlangen, FRG
As a 'subscriber' of the comp-chem mailing list you will surely have noticed
the MM[23] related traffic of the past (esp. the questions of Joe Leonard); so
one point to start for her might well be a look into the index files at the OSC
mail/ftp server.
The main point our group has been struggling with for years is the 'invention'
of new parameters for unsupported functional groups in most of our systems (or
the 'correction' of not quite realistic features, e.g. H-bonds in MM2). If
she will run into this as well, two sources to look first would be the ($1000 !)
'MM2 Forcefield Parameter Databook' discussed on the net some time ago, and a
compilation of references to published additional parameters (QCPE Bull., end
of 1991 / begin of 1992, don't have it handy right now, sorry, but I can look
it up, if needed).
Another point of importance when systems get bigger is the availability of a
easy-to-use graphics system that can display your structures (preferably
straight from the MM[23] files). Depending on the platform to run on, there
are a few programs around that combine a molecular modeling GUI with (adopted)
MM2 calcs (e.g. Hyperchem and PCModel), but there are 'graphics-only' programs
to visualise MM[23] files as well.
Implementation of the programs on systems other than a VAX is a different
problem alltogether... There are versions of MM[23] running on all possible
hardware, but as far as I know, the code you get from Allinger is still very
vaxish :-( so she may have to fiddle a bit.
I've never done crystal structure calcs myself (the way I understand the term),
but I guess if intermolecular forces are important, one would have to run a
series with increasing numbers of unit cells included in the system.
That should be it for the moment, but the probably best way to learn about
these programs is to have a go and watch everthing with a cautious mind.
2. Dr. Joe Leonard said:
Joe Leonard
jle (+ at +) world.std.com
Judging from the recent MM3 literature, if the required parameters are
present, it would be the preferred Allinger forcefield. They published
1-2 papers where they reproduce crystal structures. All in all, it looks
like a nice improvement on MM2...
3. Dr. Jim Gano mentioned:
From: Jim Gano/Chemistry <JGANO (+ at +) uoft02.utoledo.edu>
Doing calculations with just the MM2 or MM3 program is difficult
without a graphical front end. This would not be cutting edge technology
because your student would not have the file creation flexibility needed
to do good work. For a discussion of this from me see:
Gano, J.E.; Park, B.-S.; Pinkerton, A.A.; Lenoir, D.
"New Records for Sterically Congested Stilbenes:
E- and Z-1-(2,2-Dimethyl-1-tetralinylidene)-2,2-
dimethyltetralin" J. Org. Chem. 1990, 55, 2688-2693.
I suggest you get a program like PCModel. If you have some
cash and a faster computer than a PC, get Macromodel.
4. Dr. Yvonne Martin stated:
Yvonne Martin, Senior Project Leader
Computer Assisted Molecular Design Project
D-47E, AP9A-LL
Abbott Laboratories
One Abbott Park Road
IL 60064
Phone: 708 937-5362 FAX: 708 937-2625
We are big fans of MM2, haven't really evaluated MM3 yet.
For a new project we routinely find several similar molecules in the
Cambridge structural database, enough molecules to cover all the functional
groups and substructures of interest, and put them through MM2. That way
we can see how MM2 changes the structure & if it will be suitable.
We do not search conformational space with MM2, rather we use distance
geometry & minimize those structures with MM2. The dihedral driver has
some documented problems "lag" because of the way it is implemented.
You might find that you are missing several parameters. Then, its up to
you. Allinger supplies his current list upon a reprint request for parameters.
A comparison of MM2, MM3, and AM1 for some project of interest would be a
neat project for a student & useful to the rest of us if it were published.
5. Dr. Thomas Strassner told me:
Thomas Strassner
email: strasner (+ at +) organik.uni-erlangen.de
I can only tell you about MM2, cause I didn't work with MM3 yet.
There are special cards to calculate cristal structures (see manual).
The very latest parameters are available, please send me a mail if you
need them. I got them this summer of N.L. Allinger.
I hope for your friend, that the structures contain only those bonds for
which parameters are included in the program.
If not, he will have a serious problem, where to get reliable parameters
for his systems.
If you can tell me about the type of molecule he wants to calculate,
perhaps I can give him some hints where to find those.
6. Dr. Dora Schmur suggested:
Dora Schnur
head computationalchemistry/modeling
Monsanto Agricultural Group
800 N. Lindbergh Blvd mailstop U3E
St Louis, Mo 63167
phone 314-694-1563
fax 314-694-4039
dmschn (+ at +) bb1t.monsanto.com
Feng WANG
hat kind of info does your friend need on mm2/3? If s/he needs to learn the
basics of running the code, I suggest A Handbook of Computational Chemistry by
Tim Clark and also the chapter in vol2 of Reviews in Computational Chemistry
(editors: Boyd and Lipkowitz) by Allinger and Bowen.
It might be worth pointing out that MM2 and MM3 are parameterized using
electron diffraction and neutron diffraction in preference to x-ray where
possible so the results describe gas phase isolated molecules rather than
crystals. Allinger has however published papers on benzene crystal packing
so there are appropriate uses for MM2 for crystals. I reccommend contacting
Allinger himself.
Since I am an extended member of Allingers group and have done parameter
development for years(including some freeware available from the osc archives
for parameter generation), your friend may contact me as well. (But, since I
work in industry, I can't guarantee prompt responses.)
------------------------end of file---------------------------------------------