Prediction of beta-turns

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Biology

Background:

• Globular proteins contain chain reversal regions crucial for their 3D structure.
• Identifying these beta-turns is essential for understanding protein folding and function.

Purpose of the Study:

• To develop an automated computer algorithm for predicting beta-turn regions in globular proteins.
• To analyze the correlation between protein secondary structure content and beta-turn frequency.

Main Methods:

• Utilized bend frequencies and beta-turn conformational parameters (Pt) derived from X-ray crystallographic data of 408 beta-turns in 29 proteins.
• Developed a computational approach to select probable bends based on tetrapeptide probabilities (pt) and compared adjacent probable bends.
• Validated the algorithm's accuracy in predicting bend/non-bend residues and localizing beta-turns.

Main Results:

• The algorithm achieved 70% accuracy in predicting bend and non-bend residues and 78% accuracy in localizing beta-turns within +/- 2 residues.
• Helical proteins showed lower beta-turn content (17%) compared to beta-sheet proteins (41%).
• Proteins with iron-sulfur clusters, such as Chromatium high potential iron protein and rubredoxin, exhibited the highest beta-turn percentages (65%).

Conclusions:

• The automated prediction method effectively identifies beta-turn regions in globular proteins.
• Beta-turn frequency varies significantly with protein secondary structure composition.
• The study provides a valuable tool for protein structure prediction and analysis.