Weighted persistent homology for osmolyte molecular aggregation and hydrogen-bonding network analysis

Area of Science:

• Computational Chemistry
• Biophysics
• Materials Science

Background:

• Trimethylamine N-oxide (TMAO) and urea are osmolytes with known but incompletely understood differential effects on protein folding.
• Existing theoretical and experimental studies have not fully elucidated the molecular mechanisms behind these differences.
• Understanding these mechanisms is crucial for protein stability and function research.

Purpose of the Study:

• To systematically investigate the molecular aggregation and hydrogen-bonding networks of TMAO and urea using weighted persistent homology.
• To analyze the local topological structures and interactions of these osmolytes.
• To provide a novel quantitative method for characterizing molecular systems.

Main Methods:

• Application of weighted persistent homology, specifically localized persistent homology (LPH) and interactive persistent homology (IPH) models.
• Quantitative characterization of topological features using Boltzmann persistent entropy (BPE) and persistent Betti number (PBN).
• Development and application of persistent radial distribution functions (PRDFs) from IPH for analyzing molecular interactions.

Main Results:

• TMAO exhibits a distinct local network structure with increasing concentration, characterized by more numerous but smaller circle elements.
• Urea displays two local topological patterns: clusters around 6 Å and global circle elements near 12 Å.
• IPH-derived PRDFs correlate with traditional radial distribution functions and reveal significant differences in TMAO and urea interactions between 5-10 Å filtration sizes.

Conclusions:

• Weighted persistent homology provides unprecedented insights into the localized topological differences between TMAO and urea.
• The distinct topological signatures of TMAO and urea correlate with their known differential effects on protein folding.
• This topological approach offers a versatile tool for analyzing diverse molecular systems and networks.