iRNAD

iRNAD predicts dihydrouridine (D) modification sites in RNA sequences to enable computational identification of post-transcriptional D sites across eukaryotes, bacteria, and some archaea.

Key Features:

• Prediction target: Identifies dihydrouridine (D) modification sites within RNA sequences.
• Sequence encoding: Encodes RNA samples based on nucleotide chemical properties and nucleotide density.
• Species coverage: Processes RNA samples from five different species.
• Classifier: Uses a Support Vector Machine (SVM) for classification of modification sites.
• Validation: Evaluated by jackknife cross-validation with overall accuracy of 96.18% and AUC-ROC of 0.9839.
• Biological scope: Targets dihydrouridine, a post-transcriptional modification found in eukaryotes, bacteria, and some archaea, with reported elevated levels in cancerous tissues.

Scientific Applications:

• Modification site mapping: Identification of dihydrouridine sites in RNA sequences to support mapping of post-transcriptional modifications.
• Functional studies: Facilitates investigation of dihydrouridine roles in RNA structure and nucleotide conformational flexibility.
• Comparative analysis: Enables cross-species comparison of dihydrouridine distribution among the five studied species and broader taxa.
• Cancer research: Supports analysis of altered dihydrouridine levels in cancerous tissues.

Methodology:

RNA samples from five species were encoded using nucleotide chemical properties and nucleotide density, classified with a Support Vector Machine (SVM), and evaluated by jackknife cross-validation (overall accuracy 96.18%, AUC-ROC 0.9839).