Quantifying Changes in Intrinsic Molecular Motion Using Support Vector Machines.

Area of Science:

• Molecular dynamics
• Biophysics
• Computational biology

Background:

• Molecular motion is crucial for biological function and can be altered by environmental factors or molecular binding.
• Quantifying these motion changes is vital for understanding thermodynamic shifts but challenging due to high-dimensional data.
• Existing methods reduce dimensionality before comparison, risking information loss and bias.

Purpose of the Study:

• To introduce a novel method for directly comparing molecular ensembles and quantifying induced changes in intrinsic motion.
• To apply this method to identify specific regions of the paramyxovirus G protein affected by Ephrin B2 binding.

Main Methods:

• Utilized a support vector machine-based approach for direct comparison of molecular configuration ensembles.
• Generated molecular dynamics simulations for the paramyxovirus G protein in the presence and absence of Ephrin B2.
• Quantified motion differences for each residue and mapped significant changes onto the protein's 3-D structure.

Main Results:

• Developed a normalized quantitative estimate for differences between molecular ensembles.
• Identified specific regions on the paramyxovirus G protein exhibiting altered motion upon Ephrin B2 binding.
• Observed that the most affected residues are clustered on a contiguous protein facet crucial for viral fusion.

Conclusions:

• The new method accurately quantifies changes in molecular motion without prior dimensionality reduction, minimizing bias.
• The findings highlight a specific functional facet of the paramyxovirus G protein involved in Ephrin B2-mediated viral fusion.
• This approach offers a powerful tool for analyzing molecular interactions and dynamics across various systems.