iPReditor-CMG: Improving a predictive RNA editor for crop mitochondrial genomes using genomic sequence features and...

Area of Science:

• Plant Molecular Biology
• Genomics
• Bioinformatics

Background:

• RNA editing is a crucial post-transcriptional process in plant mitochondria, converting cytidine (C) to uridine (U) in protein-coding genes.
• Efficiently identifying all C-to-U editing sites on a genomic scale is challenging, necessitating accurate prediction methods to reduce experimental effort.

Purpose of the Study:

• To develop and validate a novel computational method, iPReditor-CMG, for predicting RNA editing sites in crop mitochondrial genomes.
• To improve the efficiency and accuracy of detecting cytidine-to-uridine (C-to-U) editing events.

Main Methods:

• Genome sequences of Arabidopsis thaliana, Brassica napus, and Oryza sativa were used as training and test sets.
• An optimized support vector machine (SVM) was employed for feature selection and modeling after coding genome sequences into numerical vectors.
• Validation was performed using tobacco mitochondrial ATPase genes and transcripts.

Main Results:

• The iPReditor-CMG achieved an average intraspecific prediction accuracy of 0.85, reaching 0.91 in A. thaliana, outperforming reference models.
• Interspecific prediction accuracy was 0.78 between dicotyledons and 0.56 between dicotyledons and monocotyledons, suggesting conserved editing mechanisms in related species.
• The best model achieved 0.91 accuracy and 0.88 AUC, identifying five novel feature sequences associated with RNA editing.

Conclusions:

• iPReditor-CMG is an effective tool for predicting RNA editing sites in crop mitochondria.
• The findings contribute to a better understanding of RNA editing mechanisms and post-transcriptional regulation in plant mitochondria.
• Identified feature sequences may offer new insights into the molecular basis of C-to-U editing.