Using a hybrid neural network architecture for DNA sequence representation: A study on N4-methylcytosine sites

Van-Nui Nguyen, Trang-Thi Ho, Thu-Dung Doan, Nguyen-Quoc-Khanh Le

Abstract: Highlights•Our predictive models surpass existing approaches in identifying 4mC sites in Fragaria vesca, achieving superior accuracy and sensitivity.•Advanced word embedding techniques like fastText and variable k-mers capture key patterns crucial for accurate 4mC site prediction.•Our models uncover distinct sequence motifs at 4mC sites, highlighting key epigenetic mechanisms in gene regulation and adaptation.•These models support crop breeding strategies, environmental adaptation studies, and disease research by precisely identifying epigenetic traits.