Normal-mode driven exploration of protein domain motions

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Protein functional motions often involve a few low-frequency normal modes.
• Predicting these motions is challenging without experimental data.

Purpose of the Study:

• To develop a novel approach for accurately generating protein conformers using normal coordinates.
• To transform protein conformational change prediction into a distance-geometry problem.

Main Methods:

• A two-step method: first, generate conformers using normal coordinates to define distance constraints.
• Second, build final conformers matching these constraints using the ROSETTA software.
• Solving the distance-geometry problem with ROSETTA.

Main Results:

• Successfully rebuilt six large-amplitude protein conformational changes.
• Utilized a maximum of six low-frequency normal coordinates.
• Demonstrated accurate generation of low-energy conformers for specific proteins (LAO binding protein, lysozyme T4, adenylate kinase).

Conclusions:

• The proposed approach effectively predicts protein conformational changes.
• The method transforms normal coordinate-based prediction into a solvable distance-geometry problem.
• Low-dimensionality of the motion subspace allows for efficient conformer generation.