Measuring Intrinsic Disorder and Tracking Conformational Transitions Using Rosetta ResidueDisorder

Area of Science:

• Computational biology
• Structural bioinformatics
• Protein dynamics

Background:

• Proteins can contain intrinsically disordered regions lacking stable structures.
• Computational methods predict disorder from amino acid sequences.
• Measuring disorder directly from 3D structures is currently limited.

Purpose of the Study:

• To extend Rosetta ResidueDisorder for measuring intrinsic disorder directly from protein structures.
• To assess the accuracy of structure-based disorder prediction.
• To analyze protein folding and unfolding transitions.

Main Methods:

• Extended Rosetta ResidueDisorder to utilize protein coordinate data.
• Validated the method on a benchmark set of 229 proteins.
• Analyzed molecular dynamics (MD) simulations of protein unfolding trajectories.

Main Results:

• Structure-based disorder prediction achieved 69.2% accuracy, outperforming sequence-based prediction (65.4%).
• Observed a strong correlation between RMSD and the fraction of denatured residues during unfolding.
• Developed methods to predict folding/unfolding transitions from MD trajectories.

Conclusions:

• Intrinsic disorder can be accurately measured from single protein structures.
• The enhanced Rosetta ResidueDisorder aids in analyzing protein folding dynamics and transitions.
• The tool facilitates the study of both intrinsically disordered and non-disordered proteins.