diffBUM-HMM

diffBUM-HMM applies a noise-aware Hidden Markov Model to detect differential RNA structure from high-throughput structure-probing data produced by next-generation sequencing (NGS).

Key Features:

• Noise-Aware Modeling: Incorporates a noise-aware statistical framework to distinguish true structural signals from experimental noise in structure-probing data.
• High Sensitivity and Specificity: Enhances detection of subtle structural variations while controlling false positives.
• Compatibility Across Probing Chemistries: Supports analysis of data from a wide range of RNA structure-probing chemistries.
• Accounting for Sampling Variability: Models sampling variation and sequence coverage biases common in high-throughput sequencing datasets.

Scientific Applications:

• RNA Flexibility Analysis: Quantitatively assess RNA structural flexibility across conditions.
• Conformational Change Detection: Detect conformational shifts in RNA that may reflect functional or regulatory changes.
• Protein-RNA Interaction Studies: Map protein-RNA binding sites from differential structure-probing signals.

Methodology:

Implements a Hidden Markov Model (HMM) framework with noise-aware statistical modeling that accounts for sampling variation and sequence coverage biases.