FRET-guided selection of RNA 3D structures

Area of Science:

• Biomolecular modeling
• Structural biology
• RNA biophysics

Background:

• Predicting RNA structures is challenging due to complex energy landscapes and conformational diversity.
• Accurate RNA structural collections are crucial for understanding binding and folding.
• Existing computational methods struggle with the heterogeneity of large RNA molecules.

Purpose of the Study:

• To develop and validate a Förster resonance energy transfer (FRET)-guided strategy for predicting RNA 3D structures.
• To identify RNA conformational states consistent with single-molecule FRET (smFRET) experimental data.
• To integrate computational RNA modeling with experimental biophysical techniques.

Main Methods:

• Predicted 3D RNA structures using RNAComposer, FARFAR2, and AlphaFold3.
• Validated models based on Watson-Crick base-pairing and an eRMSD threshold.
• Computed dye pair accessible contact volumes using FRETraj to predict FRET distributions.
• Compared predicted FRET distributions with experimental smFRET data to identify compatible states.

Main Results:

• Successfully predicted RNA 3D structures that are consistent with experimental smFRET data.
• Demonstrated that in silico predicted RNA structures can reproduce experimental transfer efficiencies.
• Identified specific RNA conformational states compatible with observed FRET states.
• Validated the utility of a FRET-guided workflow for analyzing flexible RNA motifs.

Conclusions:

• The FRET-guided workflow enables accurate prediction of RNA conformational states.
• This integrative approach enhances the study of RNA folding and dynamics.
• The method provides a foundation for capturing diverse conformational states in flexible RNA motifs.