new minimizers
2 messages in this thread
A good book which discusses all types of computational methods seems to be
'Reviews in Computational Chemistry' Kenny B. Lipkowitz and Donald B. Boyd
eds. In increasing order of computer time one has molecular mechanics
(MM2, ball and spring approximation), semi-empirical (AM2, a real ? QM
Schroedinger type solution except only for bonding orbitals and with varin
[C(damn typos) with varying amounts of electron overlap and off diagonal
elements but with the empirical parameters fit to experimental data) and
then ab initio calculations (6-31G** etc, which in principle iteratively
solve the 'full' Schroedinger equation. I have a sneaky feeling some do a
much better job than others, but have not read enough here. I have seen
(somewhere) data suggesting that ab initio calculations start at 1000 fold
slower than molecular mechanics calculations. One cannot use such methods
for proteins… even on next years Cray. However, I had in mind
glucose…….
John,
Interesting. I need to read up on this stuff; my personal chemistry
background is more of the "mix it up and see what you got" style. Of
course, that's why I deal with computers now <g>. I may pick up the
Lipkowitz book you mentioned.
>> I have a sneaky feeling some do a much better job than others
This is called product differentiation. But of course, no mktg person
would ever say "better job", perhaps a more appropriate phrase would be "a
more thorough implementation of the equation"?
>> (damn typos)
I strongly recommend CIM or TAPCIS, which is what most of us use. It lets
you log on, mark threads, download them, read and reply off-line, then
upload your replies. The CIS pricing structure is really designed toward
these areas; I am active in a number of the computer forums and I can read
all the msgs while staying in two digits on my monthly bill.
It makes lurking affordable. And then you have no excuse for your typos,
'cause you compose msgs off-line! <g>
Rich