surface burial in energy calc's

John Kuszewski johnk at spasm.niddk.nih.gov
Mon Sep 19 13:48:27 EST 1994


You should also have a look at Privalov and Makhatadze's
papers last year (JMB 232. 639-659 and 660-779).  There
is also another set of parameters from Murphy and Freire,
but I don't have the reference in front of me (some 
help, Kip?) that appeared recently.

One major problem with comparing different solvation 
parameters is that they model different thermodynamic
processes:  eg., Murphy and Freire model the process of
moving from protein interior to exposed sites, but 
Privalov and Makhatadze model the process of moving 
from gas to solvated liquid.  Thus, M&F's solvation
parameters include energies such as making/breaking 
van der Waals contacts, which P&M's model treats 
separately.

Another problem is that different groups do their 
analysis with different atom "classes":  Eisenberg
et al. lump all carbons together (if I'm remembering
correctly), but P&M treat aliphatic and aromatic 
carbons separately.

Werner Braun et al. (JMB 233. 275-292, 1993) published
a paper that compares various solvation parameters 
during torsion-angle-space energy minimization.  Ooi's, 
Eisenberg's, and Scheraga's parameter sets were among 
those examined.  None worked at all well, as the 
gentleman from UCLA pointed out.

If you want to learn about solvation in general,
have a look at Ben-Naim's paper in Curr Op Struct
Biol 4. 264-268 (1994).


-- 
                                   _____________
                                   |        ___/_
                                   |        |/  /
                                   --  /\  //  /--
                                   ||  ||  /  /||
                                   ||  || /  / ||
                                   ||  ||/  /  ||
John Kuszewski                     ||  |/  /|  ||      
johnk at spasm.niddk.nih.gov          ||  /  /||  ||
                                   \/ /  / ||  \/
that's MISTER protein G to you!     |/__/|      |
                                      /_________|



More information about the Molmodel mailing list