Evaluating Geometric Definitions of Stacking for RNA Dinucleoside Monophosphates Using Molecular Mechanics...

Area of Science:

• Computational chemistry and biophysics
• Structural biology
• Pharmacology

Background:

• Small molecule RNA modulation is a novel therapeutic strategy.
• Accurate RNA structural predictions are crucial for drug development.
• Computational models using force fields offer insights into RNA dynamics.

Purpose of the Study:

• To benchmark four stacking definitions for RNA dinucleoside monophosphates (DNMPs).
• To evaluate the performance of two RNA force fields (RNA-IL and ff99OL3).
• To identify improved methods for predicting RNA stacking interactions and free energies.

Main Methods:

• Molecular dynamics (MD) simulations of 16 RNA DNMPs.
• Benchmarking four different stacking definitions.
• Comparison of predicted stacking free energies with experimental thermodynamics data.
• Discrete Path Sampling (DPS) calculations.

Main Results:

• A proposed stacking definition shows improved correlation with experimental thermodynamics data.
• Predictions achieved accuracy within 0.2 kcal/mol at 300 K for specific RNA sequences.
• Purine-pyrimidine and pyrimidine-purine DNMP stacking were generally under- or overpredicted.
• Both force fields poorly predicted population distributions for RNA UU DNMPs, indicating a need for revisions.

Conclusions:

• The developed stacking definition enhances the accuracy of computational RNA structure prediction.
• Experimental thermodynamics data serve as valuable benchmarks for testing and refining RNA force fields.
• Further force field development is necessary for accurate prediction of RNA dynamics and interactions.