Jumping between protein conformers using normal modes.

Area of Science:

• Protein dynamics and structural biology.
• Computational biophysics and molecular modeling.

Background:

• The link between protein normal modes and functional conformational changes is long-studied but challenging to apply predictively.
• Existing methods struggle to accurately model the full range of protein conformational changes.

Purpose of the Study:

• To develop a method for generating accurate protein conformers representing functional conformational changes in a single step.
• To assess the efficacy of Elastic Network Models (ENMs) in predicting these conformational changes.

Main Methods:

• Utilizing Elastic Network Models (ENMs) to calculate protein normal modes.
• Generating model conformers by following the lowest-frequency or robust modes.
• Assessing model conformer quality using ROSETTA scoring and PROCHECK.

Main Results:

• Accurate model conformers (Cα-RMSD < 1 Å) of known conformational endpoints were generated in a single step.
• The 50 lowest-frequency modes of an all-atom ENM yielded the most accurate conformers.
• Fewer than ten robust modes captured ~90% of the motion described by the 50 lowest-frequency modes.

Conclusions:

• Exploring robust modes of ENMs is an efficient strategy for sampling functionally relevant protein conformational changes.
• This approach offers a powerful tool for predicting and understanding protein function.
• The method shows promise for applications in drug discovery and protein design.