Improvement of comparative model accuracy by free-energy optimization along principal components of natural structural variation - PubMed (original) (raw)

Comparative Study

. 2004 Oct 26;101(43):15346-51.

doi: 10.1073/pnas.0404703101. Epub 2004 Oct 18.

Affiliations

Comparative Study

Improvement of comparative model accuracy by free-energy optimization along principal components of natural structural variation

Bin Qian et al. Proc Natl Acad Sci U S A. 2004.

Abstract

Accurate high-resolution refinement of protein structure models is a formidable challenge because of the delicate balance of forces in the native state, the difficulty in sampling the very large number of alternative tightly packed conformations, and the inaccuracies in current force fields. Indeed, energy-based refinement of comparative models generally leads to degradation rather than improvement in model quality, and, hence, most current comparative modeling procedures omit physically based refinement. However, despite their inaccuracies, current force fields do contain information that is orthogonal to the evolutionary information on which comparative models are based, and, hence, refinement might be able to improve comparative models if the space that is sampled is restricted sufficiently so that false attractors are avoided. Here, we use the principal components of the variation of backbone structures within a homologous family to define a small number of evolutionarily favored sampling directions and show that model quality can be improved by energy-based optimization along these directions.

PubMed Disclaimer

Figures

Fig. 1.

Fig. 1.

Variation in structure family and variation represented by PCs. (a) Superposition of the cores of all eight structures in structure family a.133.1.1. (b–d) Backbone conformations sampled along the direction defined by the first PC (b), the second PC (c), and the third PC (d).

Fig. 2.

Fig. 2.

Fraction of structural variation explained by PCs. Each line represents a structure family. With additional PCs, more and more variation can be described. The dashed line indicates that the first three PCs can describe at least 65% of the variation observed in each structure family.

Fig. 3.

Fig. 3.

Improvement of protein backbone core region by sampling along the directions defined by the first three PCs and selecting low-energy decoys by using the Rosetta energy function. rmsd improvement of the lowest-energy decoys in tests I (a), II (b), and III (c) is shown. The improvement of rmsd is measured by (rmsd of starting model - rmsd of refined model). Positive values indicate improvement of backbone conformations.

Fig. 4.

Fig. 4.

Example of successful model refinement. Red, model structure; blue, native structure; green, refined structure. The rmsd between the model structure and native structure is 2.36 Å, and the rmsd between the refined structure and native structure is 1.42 Å.

Similar articles

Cited by

References

    1. Sanchez, R., Pieper, U., Melo, F., Eswar, N., Marti-Renom, M. A., Madhusudhan, M. S., Mirkovic, N. & Sali, A. (2000) Nat. Struct. Biol. 7, Suppl., 986-990. - PubMed
    1. Baker, D. & Sali, A. (2001) Science 294, 93-96. - PubMed
    1. Stevens, R. C., Yokoyama, S. & Wilson, I. A. (2001) Science 294, 89-92. - PubMed
    1. Fu, Z., Aronoff-Spencer, E., Backer, J. M. & Gerfen, G. J. (2003) Proc. Natl. Acad. Sci. USA 100, 3275-3280. - PMC - PubMed
    1. Ogawa, H. & Toyoshima, C. (2002) Proc. Natl. Acad. Sci. USA 99, 15977-15982. - PMC - PubMed

Publication types

MeSH terms

LinkOut - more resources