Mr. Lee et al., 2.1 and 1.8 angstrom average C-alpha RMSD structure predictions on two small proteins, HP-36 and S15, J AM CHEM S, 123(6), 2001, pp. 1040-1046
On two different small proteins, the 36-mer villin headpiece domain (HP-36)
and the 65-mer structured region of ribosomal protein (S15), several model
predictions from the ab initio approach Rosetta were subjected to molecula
r dynamics simulations for refinement. After clustering the resulting traje
ctories into conformational families, the average molecular mechanics-Poiss
on Boltzmann/surface area (MM-PBSA) free energies and alpha carbon (C-alpha
) RMSDs were then calculated for each family. Those conformational families
with the lowest average free energies also contained the best C-alpha RMSD
structures (1.4 Angstrom for S15 and HP-36 core) and the lowest average C-
alpha RMSDs (1.8 Angstrom for S15, 2.1 Angstrom for HP-36 core). For compar
ison, control simulations starting with the two experimental structures wer
e very stable, each consisting of a single conformational family, with an a
verage C-alpha RMSD of 1.3 Angstrom for S15 and 1.2 Angstrom for HP-36 core
(1.9 Angstrom over all residues). In addition, the average free energies'
ranks (Spearman rank, r(s)) correlate well with the average C-alpha RMSDs (
r(s) = 0.77 for HP-36, r(s) = 0.83 for S15). Molecular dynamics simulations
combined with the MM-PBSA free energy function provide a potentially power
ful tool for the protein structure prediction community in allowing for bot
h high-resolution structural refinement and accurate ranking of model predi
ctions. With all of the information that genomics is now providing, this me
thodology may allow for advances in going from sequence to structure.