SHAPER

SHAPER predicts SHAPE reactivities from RNA three-dimensional (3D) structures to estimate local nucleotide flexibility and inform structure modeling.

Key Features:

• SHAPE reactivity prediction: Computes per-nucleotide SHAPE reactivities from RNA 3D structural models.
• Structural parameter analysis: Analyzes 2D base pairs, solvent accessibility of 2'-OH sites, stacking interactions between nucleotides, and ribose sugar conformations to inform predictions.
• Sequence-dependent bias correction: Incorporates sequence-dependent biases into the predictive framework to adjust estimated reactivities.
• Integration with experimental SHAPE data: Supports incorporation of experimental SHAPE data as constraints in both 2D and 3D modeling workflows.
• Multiple-conformation support: Accepts RNA structures with single or multiple conformations for analysis.
• Output and validation metrics: Produces predicted SHAPE profiles and computes correlation metrics compared to provided experimental SHAPE data.
• Model foundation: Extends the 3DSSR model by integrating additional sequence-dependent and structural features.
• Performance: Provides fast and accurate SHAPE reactivity predictions.

Scientific Applications:

• RNA structure refinement: Uses predicted SHAPE reactivities and experimental constraints to improve 2D and 3D RNA structural models.
• Local flexibility estimation: Estimates nucleotide-level flexibility to support studies of RNA function, interactions, and dynamics.
• Model validation: Enables validation of computational or experimental RNA structures by comparing predicted SHAPE profiles to experimental data.

Methodology:

Computes per-nucleotide SHAPE reactivity from RNA 3D models by analyzing 2D base pairs, 2'-OH solvent accessibility, nucleotide stacking, ribose sugar conformations, and sequence-dependent biases; supports multiple conformations, integrates experimental SHAPE data as constraints in 2D and 3D modeling, and reports correlation metrics between predicted and experimental profiles.