SoftVoting6mA: An improved ensemble-based method for predicting DNA N6-methyladenine sites in cross-species genomes

Area of Science:

• Epigenetics
• Genomics
• Bioinformatics

Background:

• DNA N6-methyladenine (6mA) is a key epigenetic modification regulating gene expression, DNA replication, repair, and recombination.
• Accurate identification of 6mA sites is essential for understanding its functional significance but remains a challenge.

Purpose of the Study:

• To develop an improved, cross-species method for predicting DNA N6-methyladenine sites.
• To enhance the accessibility of 6mA site prediction tools through a web server.

Main Methods:

• Proposed SoftVoting6mA, an ensemble-based prediction method.
• Utilized four encoding methods (electron-ion-interaction pseudo potential, One-hot encoding, Kmer, pseudo dinucleotide composition) selected for optimal performance.
• Integrated four learning algorithms using a soft voting strategy.

Main Results:

• SoftVoting6mA demonstrated state-of-the-art performance in predicting 6mA sites.
• The method was validated using 5-fold cross-validation and independent tests.
• A user-friendly web server was developed for public access.

Conclusions:

• SoftVoting6mA offers a highly effective and accessible solution for predicting DNA N6-methyladenine sites.
• The developed tool facilitates further research into the functional roles of 6mA in various genomes.