GlyinsRNA

GlyinsRNA predicts glycosylation sites on small RNAs to identify RNA glycosylation positions that may modulate post-transcriptional gene regulation.

Key Features:

• Machine Learning-Based Prediction: Employs machine learning models using multiple RNA sequence representation encodings to capture sequence patterns surrounding glycosylation sites.
• Performance Metrics: Demonstrates AUROC of 0.7933 in five-fold cross-validation and AUROC of 0.7979 on independent testing.
• RBP Motif Annotation: Annotates predicted glycosylation sites with overrepresented RNA-binding protein (RBP)-related motifs to link sites to potential RBP interactions.

Scientific Applications:

• Post-transcriptional regulation studies: Identification of candidate glycosylation sites on small RNAs to support investigations of post-transcriptional gene regulation.
• Functional analysis of RNA modifications: Generation of hypotheses about how RNA glycosylation may affect molecular mechanisms of gene expression regulation.
• RBP interaction inference: Prioritization of sites for experimental follow-up by linking predicted glycosylation sites to overrepresented RBP-related motifs.

Methodology:

Machine learning models trained on multiple RNA sequence representation encodings, evaluated by five-fold cross-validation and independent testing (AUROC 0.7933 and 0.7979 respectively), with predicted sites annotated for overrepresented RBP-related motifs.