dsRID

dsRID identifies double-stranded RNA (dsRNA) regions from long-read RNA sequencing data to characterize the dsRNAome for studies of innate immunity and human disease.

Key Features:

• Machine Learning-Based Prediction: dsRID employs machine learning algorithms to predict dsRNA regions from sequencing-derived features with high accuracy.
• Long-Read RNA-seq (PacBio) Utilization: The method operates on PacBio long-read RNA-seq data to capture comprehensive sequence and structural information relevant to dsRNA detection.
• Training on Alzheimer's Disease Brain Samples: Models are trained using long-read RNA-seq data derived from Alzheimer's disease (AD) brain samples.
• Validation Across Multiple Datasets: The models have been validated across multiple datasets to assess robustness and accuracy.
• Cohort-Level Application: The approach has been applied to AD and control cohorts to identify potential differences in dsRNA profiles.

Scientific Applications:

• Innate Immune Response Studies: Identifying dsRNA regions supports investigation of dsRNA-triggered innate immune activation mechanisms.
• Disease Mechanism Elucidation: Characterizing global dsRNA profiles in Alzheimer's disease enables exploration of links between dsRNA expression patterns and disease pathology.
• Comparative Analysis: Comparing dsRNA profiles across cohorts, including ENCODE-sequenced samples, facilitates identification of condition-associated expression patterns.

Methodology:

Machine learning models are trained on PacBio long-read RNA-seq data from Alzheimer's disease brain samples, validated across multiple datasets, and applied to AD and control cohorts to detect and compare predicted dsRNA regions.