TargetM6A: Identifying N⁶-Methyladenosine Sites From RNA Sequences via Position-Specific Nucleotide Propensities and...

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• N6-methyladenosine (m6A) is a crucial RNA modification involved in essential biological processes.
• Accurate identification of m6A sites is vital for biomedical research and therapeutic development.

Purpose of the Study:

• To develop a computational method, TargetM6A, for rapid and accurate prediction of m6A sites from RNA sequences.
• To introduce and integrate novel features for enhanced sequence representation.

Main Methods:

• Developed TargetM6A, a computational tool utilizing position-specific nucleotide/dinucleotide propensities (PSNP/PSDP) and nucleotide composition (NC) features.
• Employed incremental feature selection (IFS) for feature subset optimization.
• Trained a support vector machine (SVM) prediction engine on the optimized feature set.

Main Results:

• TargetM6A demonstrated superior performance compared to existing methods in predicting m6A sites.
• Achieved high prediction accuracy with a Matthews correlation coefficient (MCC) of 0.526 and an area under the curve (AUC) of 0.818.
• Experimental validation included stringent jackknife and independent tests.

Conclusions:

• TargetM6A offers a robust and efficient approach for m6A site identification.
• The developed method significantly advances the prediction accuracy of m6A sites.
• A publicly accessible web server is available for TargetM6A at http://csbio.njust.edu.cn/bioinf/TargetM6A.