Topological methods for exploring low-density states in biomolecular folding pathways.

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Biophysics

Background:

• Characterizing transient intermediate or transition states is vital for understanding biomolecular folding pathways.
• These low-population states are challenging to detect in experiments and computer simulations.
• Debates exist regarding the number of intermediate states, even for simple systems like RNA hairpins.

Purpose of the Study:

• To develop a computational approach for exploring low-populated transition or intermediate states in biomolecular folding.
• To apply this method to analyze the folding pathways of RNA hairpins.

Main Methods:

• Utilized a topological data analysis tool, MAPPER, with large-scale distributed computing simulation data.
• Employed a conditional density filter to focus on pathway structures.
• Applied clustering analysis on level sets to differentiate low-populated intermediates from high-populated states.

Main Results:

• Successfully identified and characterized multiple intermediate states in the folding/unfolding pathways of an RNA hairpin with a GCAA tetraloop.
• Provided structural evidence from computer simulations supporting the existence of these intermediate states.
• Demonstrated the method's effectiveness in handling heterogeneous distributions and capturing features across multiple pathways.

Conclusions:

• The developed computational approach, inspired by Morse theory, systematically explores low-density intermediate states in complex biomolecular folding.
• The method is robust, less sensitive to distance metrics than other techniques, and adaptable for extensions.
• Offers a powerful tool for advancing the understanding of molecular folding dynamics and identifying transient states.