Details of Single-Molecule Force Spectroscopy Data Decoded by a Network-Based Automatic Clustering Algorithm.

Area of Science:

• Biophysics
• Materials Science
• Biotechnology

Background:

• Atomic force microscopy-single-molecule force spectroscopy (AFM-SMFS) is crucial for studying biological interactions under physiological conditions.
• Current analysis algorithms for AFM-SMFS data struggle to link unbinding forces to specific binding sites due to limitations in modeling relationships.
• There is a need for advanced algorithms to precisely analyze the large datasets generated by AFM-SMFS.

Purpose of the Study:

• To develop an unsupervised, network-based algorithm for analyzing AFM-SMFS data.
• To enable the precise identification of unbinding forces associated with individual binding sites.
• To overcome the limitations of existing algorithms in decoding molecular interaction details.

Main Methods:

• Development of a network-based automatic clustering algorithm (NASA).
• Application of NASA to analyze AFM-SMFS curves from heparan sulfate (HS)-antithrombin (AT) interactions on endothelial cell surfaces.
• Utilizing NASA for unsupervised clustering of unbinding force data.

Main Results:

• NASA successfully detects peaks and calculates unbinding forces from AFM-SMFS curves.
• The algorithm automatically identified three distinct unbinding force clusters for HS-AT interactions on different cell lines (Ext1^f/f and Ndst1^f/f).
• These clusters suggest the presence of three different binding sites, providing detailed molecular interaction insights.

Conclusions:

• NASA is an effective unsupervised method for decoding specific molecular details, such as unbinding forces per binding site, from AFM-SMFS data.
• The algorithm demonstrates significant potential for application in various AFM-based SMFS measurements involving "saw-tooth" force-distance curves.
• NASA offers a powerful tool for advancing the analysis of molecular interactions in fields like immunology and genetics.