LigandRNA: computational predictor of RNA-ligand interactions

Area of Science:

• Biochemistry
• Computational Biology
• Drug Discovery

Background:

• RNA molecules are increasingly recognized as crucial drug targets.
• Experimentally determined RNA 3D structures facilitate structure-based drug design.
• Computational methods for RNA-ligand docking are less developed than protein-ligand methods.

Purpose of the Study:

• To develop LigandRNA, a scoring function for predicting RNA-small molecule interactions.
• To provide a computational tool for structure-based RNA drug discovery.

Main Methods:

• Developed LigandRNA, a scoring function using a grid-based algorithm.
• Utilized a knowledge-based potential derived from experimental RNA-ligand complexes.
• Input: RNA receptor and ligand pose files. Output: Ranked ligand poses.

Main Results:

• LigandRNA accurately predicts RNA-small molecule interactions.
• LigandRNA outperforms five other publicly available docking methods.
• Combining LigandRNA with Dock6 (meta-predictor) further enhances pose prediction.

Conclusions:

• LigandRNA is an effective tool for predicting RNA-ligand binding poses.
• The LigandRNA web server is freely available for research.
• This tool can advance structure-based drug design targeting RNA molecules.