cgRNASP

cgRNASP evaluates RNA three-dimensional structures using coarse-grained, residue-separation-based knowledge-derived statistical potentials.

Key Features:

• Residue-Separation-Based Potentials: Separates interactions by residue separation into long-ranged and short-ranged terms for scoring RNA structures.
• Multiple Coarse-Grained Levels: Implements statistical potentials at various coarse-grained (CG) levels to evaluate both CG and all-atom RNA structures.
• Knowledge-Based Statistical Potentials: Derives scoring functions from observed structural data to capture RNA conformational preferences.
• Performance and Efficiency: Shows performance on par with or slightly better than rsRNASP on extensive test datasets and is more computationally efficient than traditional all-atom statistical potentials and scoring functions.
• Evaluation on Realistic Datasets: Demonstrates improved performance on realistic benchmark datasets, including RNA-Puzzles.

Scientific Applications:

• RNA Structure Prediction and Evaluation: Scores and discriminates RNA 3D conformations to support prediction and model selection.
• High-Throughput Computational Analysis: Enables large-scale evaluation of RNA structures due to reduced computational cost relative to all-atom potentials.

Methodology:

Derives knowledge-based statistical potentials at multiple coarse-grained levels and partitions interactions by residue separation into short-ranged and long-ranged terms to score RNA three-dimensional structures.