Multiscale modeling of macromolecular conformational changes combining concepts from rigidity and elastic network...

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Macromolecular conformational changes are crucial for biological function.
• Accurate modeling of these changes requires efficient computational methods.
• Existing methods may struggle with large-scale protein dynamics.

Purpose of the Study:

• To develop a novel two-step multiscale modeling approach for macromolecular conformational changes.
• To improve the efficiency and accuracy of predicting protein dynamics.
• To leverage recent advancements in rigidity and elastic network theory.

Main Methods:

• A two-step approach combining graph-theoretical decomposition (FIRST) and rotations-translations of blocks (RTB) with elastic network models.
• Step 1: Decomposes macromolecules into rigid clusters and identifies flexible links using an all-atom representation.
• Step 2: Models dynamics using coarse-grained representations where rigid clusters undergo rigid body motions.

Main Results:

• The approach significantly reduces memory requirements (avg. 9x) and computational time (avg. 27x) through coarse-graining.
• Predicted motion directions and magnitudes align well with experimental data.
• The method shows superior performance for movements dominated by loop or fragment motions.

Conclusions:

• Explicitly distinguishing between rigid and flexible regions is advantageous for simplified protein representations.
• The multiscale approach accurately predicts both large-scale motions and atomic movements within rigid clusters.
• This method offers a powerful tool for understanding protein dynamics and correlated motions.