Nearest neighbor rules for RNA helix folding thermodynamics: improved end effects

Area of Science:

• Biochemistry
• Molecular Biology
• Computational Biology

Background:

• Nearest neighbor parameters are widely used for RNA folding stability estimation.
• Current models penalize terminal AU pairs more than GC pairs.
• Terminal penalties in existing models lack sequence-context dependence.

Purpose of the Study:

• To develop refined nearest neighbor parameters for RNA secondary structure stability.
• To incorporate sequence-dependent terminal effects into RNA folding models.
• To improve the accuracy of predicting RNA helix stability.

Main Methods:

• Curated an expanded database of RNA helix stabilities from optical melting experiments.
• Analyzed updated database to identify sequence-dependent terminal penalties.
• Developed and tested new nearest neighbor parameters incorporating sequence context.

Main Results:

• Terminal penalties are dependent on the identity of the adjacent penultimate base pair.
• New parameters accurately predict helix stabilities (R=0.982) for 271 helices.
• Improved prediction accuracy for duplexes with terminal AU and GU pairs (avg. absolute difference reduced from 1.38 to 0.27 kcal/mol).

Conclusions:

• Refined understanding of helix end stability is essential for accurate RNA folding predictions.
• Sequence-dependent nearest neighbor parameters significantly enhance stability estimation.
• The updated model eliminates the need for separate treatment of certain GU wobble pairs.