The dataset for protein-RNA binding affinity.

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Protein-RNA interactions are crucial for numerous biological processes.
• Accurate prediction of binding affinity is essential for understanding and engineering these interactions.
• Existing benchmark datasets and scoring functions for protein-RNA docking require further development.

Purpose of the Study:

• To create a non-redundant benchmark dataset of protein-RNA complexes with experimentally measured binding affinities.
• To evaluate the performance of current protein-RNA scoring functions.
• To provide a resource for the development and comparison of new scoring functions.

Main Methods:

• Compiled a dataset of 73 protein-RNA complexes from the Protein Data Bank.
• Included experimentally measured binding affinity data (Kd) and associated conditions (pH, temperature).
• Calculated binding free energy from Kd and compared it with predictions from three established protein-RNA docking scoring functions.

Main Results:

• The developed dataset comprises 73 protein-RNA complexes with binding affinity data.
• A low correlation was observed between predicted binding free energy from scoring functions and experimentally derived binding Gibbs free energy.
• This indicates limitations in the current scoring functions for accurately predicting protein-RNA binding strength.

Conclusions:

• The new benchmark dataset provides a valuable resource for advancing protein-RNA interaction research.
• Current scoring functions demonstrate insufficient accuracy for predicting protein-RNA binding affinity.
• Further research is needed to develop more reliable computational models for protein-RNA interactions.