Replacing salt correction factors with optimized RNA nearest-neighbour enthalpy and entropy parameters

Area of Science:

• Biophysics
• Computational Biology
• Molecular Biology

Background:

• Accurate prediction of RNA melting temperatures (Tm) is crucial for understanding RNA folding and function.
• Existing thermodynamic models often require salt corrections, adding complexity and potential error.
• Nearest-neighbor parameters are fundamental to these thermodynamic models.

Purpose of the Study:

• To develop new nearest-neighbor thermodynamic parameters for RNA.
• To evaluate the performance of these parameters with and without salt correction.
• To introduce an interpolation scheme for predicting parameters at various salt concentrations.

Main Methods:

• Calculation of nearest-neighbor enthalpies and entropies for 18 RNA sequences across 5 salt concentrations.
• Utilizing a melting temperature optimization method for 757 data points.
• Testing two parameter schemes: fixed and variable initial parameters.
• Developing an interpolation scheme for arbitrary salt concentrations.

Main Results:

• New parameters show a quadratic dependence on salt concentration, which linearizes Gibbs free energies.
• Variable initial parameters yielded superior predictive results compared to salt correction factors.
• Significantly reduced prediction uncertainty was observed for independent validation sequences.
• The interpolation scheme outperformed predictions using traditional salt corrections.

Conclusions:

• The newly derived RNA nearest-neighbor parameters provide improved melting temperature predictions.
• These parameters do not require salt correction and offer enhanced accuracy.
• The variable initial parameter scheme and interpolation method represent advancements in RNA thermodynamic modeling.