HYPROSP: a hybrid protein secondary structure prediction algorithm—a knowledge-base approach

November 1, 2004

Yi-Chiao, Fang

Abstract

Protein secondary structures serve as important building blocks in comparative modeling and protein threading methods for protein 3D structure prediction. They can be used to generate templates for protein 3D structure prediction algorithms to build protein structure models. PROSP is a knowledge-base approach for protein secondary structure prediction. The knowledge base here contains small peptide fragments together with their structural information. The match rate is the amount of structural information that a target protein can extract from the peptides in the knowledge base. Target proteins with higher match rate are likely predicted more accurately based on PROSP developed by the authors. The hybrid protein structure prediction (HYPROSP) approach is that when match rate is at least 80%, using PROSP, otherwise, using PSIPRED, to predict protein secondary structure. While the match rate is more than 80%, the average Q₃ of PROSP is 3.96 and 7.2 better than that of PSIPRED on DSSP and EVA data, respectively.

Source

Nucleic Acids Res. 2004 Sep 24;32(17):5059-65.

Kuen-Pin Wu, Hsin-Nan Lin, Jia-Ming Chang, Ting-Yi Sung and Wen-Lian Hsu.

Reference

1. Rost B. Review: protein secondary structure prediction continues to rise. J Struct Biol 134, 204-218 (2001).

2. Kabsch W. & Sander C. Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers. 22, 2577-2637 (1983).

3. Rost B. & Eyrich V. A. EVA: large-scale analysis of secondary structure prediction. Proteins Suppl 5, 192-199 (2001)

4. Jones D.T. Protein secondary structure prediction based on position-specific scoring matrices. J Mol Biol 292, 195-202 (1999).

5. Rost B. & Sander C. Third generation prediction of secondary structure. In: Webster D. M. (ed.): 'Protein structure prediction: methods and protocols'. Humana Press, 2000, 71-95.