Protein-RNA Docking Using ICM.

Area of Science:

• Computational biology
• Structural biology
• Biochemistry

Background:

• Protein-RNA interactions are crucial for numerous biological processes.
• Computational docking methods are essential for predicting complex structures.
• Existing methods require refinement for accurate protein-RNA complex modeling.

Purpose of the Study:

• To evaluate and enhance the performance of an FFT-based rigid-body docking algorithm for protein-RNA complexes.
• To improve the accuracy of computational prediction for protein-RNA structures.
• To optimize scoring functions for better identification of correct poses.

Main Methods:

• Utilized three recent protein-RNA complex datasets.
• Modified the energy function to include a specific electrostatic term for phosphate-protein interactions.
• Optimized a scoring function incorporating van der Waals, electrostatic, and solvation terms.
• Employed pose clustering using contact fingerprints.

Main Results:

• Incorporating an electrostatic term improved the docking step.
• Optimization significantly reduced the weight of the solvation term in the scoring function.
• Rescoring with the new function increased success rates.
• Pose clustering further enhanced accuracy, achieving a 0.66 success rate for top 100 structures.

Conclusions:

• The refined FFT-based docking algorithm and optimized scoring function significantly improve the prediction of protein-RNA complex structures.
• Electrostatic interactions play a key role, while solvation energy is less critical for scoring.
• The developed computational approach offers a more accurate tool for studying protein-RNA interactions.