johnk at spasm.niddk.nih.gov
Tue Aug 1 11:37:17 EST 1995
In article <3vl1dv$agf at sun4.bham.ac.uk>, s.m.williams.bcm at bham.ac.uk (Steve) writes:
|> kballi at aol.com (KBalli) wrote:
|> >I am new to the science/art of protein modeling and looking for some
|> >advice or pointers. I just started working in a cell biology lab studying
|> >gap junction proteins at Loma Linda University School of Medicine. My
|> >goal is to add conformational / energy minimization data using computer
Errum, exactly what do you mean by "energy minimization?"
In case you didn't know, the energy terms used in all molecular dynamics
programs (like CHARMm) are extremely crude, and the trajectories and
"minimum energy" conformations they produce of proteins have extremely
little to do with reality.
What structures were you intending to refine? If you're starting from
a crystal structure, it has almost certainly been "energy minimized"
already (using X-PLOR) in a way that takes into account both experimental
data (crystallography or NMR) from that structure and a priori information
(eg., bond lengths, planar groups, etc.).
If you're trying to start from a 10 Angstrom "structure," there's no
hope that optimizing any molecular dynamics "energy" would give you
anything that has anything to do with reality.
|> >Currently our lab has no software, so I am working in another lab using
|> >Quanta 4.0 / CHARMm on a Silicon Graphics IRIS 4D (circa early 80's I
|> >believe). No one else in the lab uses it and I am spending lots of time
|> >wading through their 8 manuals. The software seems quite comprehensive
|> >but I have nothing to compare it to. Our lab is hoping to purchase some
|> >software / hardware of our own soon. A couple questions I've been
|> >wondering about are:
|> >1) How does this software compare to others out on the market. Is it
|> >industry standard, or is there better?
CHARMm is very good by molecular dynamics standards, and is useful for
very small molecules in simple states (eg., ideal gasses).
|> >2) What other software is available for UNIX?
There are programs from Biosym (Discover), Tripos (whose name I can't
recall), and Molecular Simulations (CHARMm, X-PLOR). They all do
If your protein has fairly high homology to another protein whose
structure is known, then there are different programs that can
try to get an idea of what your protein's structure might look like.
The best of these is Andrej Sali's MODELER (which is also sold by
|> >3) Our lab currently has only PC's and Power Mac's - is there any
|> >comparable software made for these platforms?
Not much. Most serious modeling/molecular dynamics programs only
run under UNIX.
|> >4) A friend told me about a software package called Imdad / Look from
|> >Molecular Applications Group - is this any good?
MacImdad is just a pretty picture program. It does not try to change
the coordinates it's handed.
|> >5) Are there any good books, journals, or other Internet resources where I
|> >might find answers to these and other molecular modeling questions as I
|> >get started on this project?
For a quick, painless start in protein dynamics and energy, read
J. Andrew McCammon's "Dynamics of Proteins and Nucleic Acids" book.
It's somewhat dated, but it gets you the fundamentals quickly and
|> You could try RASMOL for the PC. I have a link on one of my pages:-
No he couldn't. Read his question. RasMol is just a pretty-picture
maker. It does no conformational analysis or "energy" minimization.
| |/ /
-- /\ // /--
|| || / /||
|| || / / ||
|| ||/ / ||
John Kuszewski || |/ /| ||
johnk at spasm.niddk.nih.gov || / /|| ||
\/ / / || \/
that's MISTER protein G to you! |/__/| |
"Biophysics has driven me to an attitude of apocalyptic doom"
More information about the Proteins