m6AmPred

m6AmPred predicts N6,2'-O-dimethyladenosine (m6Am) modification sites in RNA sequences to support analysis of their roles in mRNA regulation.

Key Features:

• eXtreme Gradient Boosting (XgbDart): Implements the XgbDart algorithm for classification and reports high predictive performance (AUC > 0.954) in 10-fold cross-validation and independent testing.
• EIIP-PseEIIP encoding: Transforms RNA primary sequences using an EIIP-PseEIIP scheme that encodes nucleotide physicochemical properties for machine learning input.
• Validation on experimental data: Trained and validated using experimentally verified m6Am sites compiled from two distinct data sources.

Scientific Applications:

• m6Am site prediction: Enables in silico identification of N6,2'-O-dimethyladenosine sites within RNA sequences.
• Post-transcriptional regulation studies: Supports investigations into effects of m6Am on mRNA stability, translation, and degradation.
• Modification mapping: Facilitates mapping of m6Am modification patterns across different RNA types and experimental conditions.

Methodology:

RNA primary sequences are encoded with EIIP-PseEIIP and input to an eXtreme Gradient Boosting with Dart (XgbDart) classifier, with performance evaluated by 10-fold cross-validation and independent testing using experimentally verified m6Am sites (AUC > 0.954).