A. Godzik et al., DE-NOVO AND INVERSE FOLDING PREDICTIONS OF PROTEIN-STRUCTURE AND DYNAMICS, Journal of computer-aided molecular design, 7(4), 1993, pp. 397-438
In the last two years, the use of simplified models has facilitated ma
jor progress in the globular protein folding problem, viz., the predic
tion of the three-dimensional (3D) structure of a globular protein fro
m its amino acid sequence. A number of groups have addressed the inver
se folding problem where one examines the compatibility of a given seq
uence with a given (and already determined) structure. A comparison of
extant inverse protein-folding algorithms is presented, and methodolo
gies for identifying sequences likely to adopt identical folding topol
ogies, even when they lack sequence homology, are described. Extension
to produce structural templates or fingerprints from idealized struct
ures is discussed, and for eight-membered beta-barrel proteins, it is
shown that idealized fingerprints constructed from simple topology dia
grams can correctly identify sequences having the appropriate topology
. Furthermore, this inverse folding algorithm is generalized to predic
t elements of supersecondary structure including beta-hairpins, helica
l hairpins and alpha/beta/alpha fragments. Then, we describe a very hi
gh coordination number lattice model that can predict the 3D structure
of a number of globular proteins de novo; i.e. using just the amino a
cid sequence. Applications to sequences designed by DeGrado and co-wor
kers [Biophys. J., 61 (1992) A265] predict folding intermediates, nati
ve states and relative stabilities in accord with experiment. The meth
odology has also been applied to the four-helix bundle designed by Ric
hardson and co-workers [Science, 249 (1990) 884] and a redesigned mono
meric version of a naturally occurring four-helix dimer, rop. Based on
comparison to the rop dimer, the simulations predict conformations wi
th rms values of 3-4 angstrom from native. Furthermore, the de novo al
gorithms can assess the stability of the folds predicted from the inve
rse algorithm, while the inverse folding algorithms can assess the qua
lity of the de novo models. Thus, the synergism of the de novo and inv
erse folding algorithm approaches provides a set of complementary tool
s that will facilitate further progress on the protein-folding problem
.