A computational pathway for bracketing native-like structures fo small alpha helical globular proteins.

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Protein Folding

Background:

• Protein structure prediction is crucial but computationally challenging, often described as an NP-hard problem.
• Existing ab initio methods face insurmountable time-scale limitations, hindering rapid prediction.
• Growing structural databases offer resources but don't fully solve the prediction dilemma.

Purpose of the Study:

• To develop a computational pathway for predicting native-like protein structures for small alpha-helical globular proteins.
• To integrate biophysical filters and computational tools to efficiently search the conformational space.
• To reduce the complexity of the protein folding problem for practical applications.

Main Methods:

• An automated protocol generating multiple protein structures from secondary structural information.
• Application of knowledge-based biophysical filters (persistence length, radius of gyration) for screening structures.
• Utilizing Monte Carlo optimizations and energy minimization with a validated scoring function.

Main Results:

• The method successfully brackets native-like structures for small alpha-helical globular proteins.
• Biophysical filters significantly reduce the library of probable protein structures.
• The final ensemble of optimized structures contains candidates within 3-5 angstroms of the native structure.

Conclusions:

• The developed computational pathway offers a promising solution for predicting protein structures.
• This approach effectively narrows down the search space, making the protein folding problem more tractable.
• The method demonstrates encouraging results on twelve small alpha-helical globular proteins, advancing structural bioinformatics.