iRNA-m2G

iRNA-m2G predicts N2-methylguanosine (m2G) sites in eukaryotic transcriptomes to identify locations of this posttranscriptional modification.

Key Features:

• Computational prediction: Uses sequence-derived information to predict m2G sites.
• Encoding methodology: Encodes RNA sequences using nucleotide chemical properties and accumulated nucleotide frequency.
• Validation methods: Validated by jackknife testing and cross-species validations on benchmark datasets.
• Performance: Demonstrates promising accuracy on benchmark and independent datasets.
• Training datasets: Trained and tested on benchmark datasets S1 (143 m2G and 143 non-m2G sequences from H. sapiens, M. musculus, and S. cerevisiae) and S2 (143 non-m2G sequences from S1 plus 1246 additional non-m2G sequences from the same species).

Scientific Applications:

• RNA modification mapping: Identification of N2-methylguanosine (m2G) sites across eukaryotic transcriptomes.
• tRNA biology studies: Supports investigation of m2G roles in tRNA function and stability.

Methodology:

Encode RNA sequences using nucleotide chemical properties and accumulated nucleotide frequency, apply the computational predictor to identify candidate m2G sites, and evaluate performance using jackknife testing and cross-species validation on benchmark datasets S1 and S2.