Precision DNA methylation typing via hierarchical clustering of Nanopore current signals and attention-based neural network

• Published in: Briefings in bioinformatics Vol. 25; no. 6
• Main Authors: Dai, Qi, Chen, Hu, Yi, Wen-Jing, Zhao, Jia-Ning, Zhang, Wei, He, Ping-An, Liu, Xiao-Qing, Zheng, Ying-Feng, Shi, Zhuo-Xing
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 23.09.2024; Oxford Publishing Limited (England)
• Subjects: Artificial neural networks; Chemical treatment; Cluster Analysis; Clustering; Decoding; Deoxyribonucleic acid; DNA; DNA - chemistry; DNA - genetics; DNA Methylation; DNA sequencing; Gene sequencing; Humans; Nanopore Sequencing - methods; Nanopores; Neural coding; Neural networks; Neural Networks, Computer; Nucleotide sequence; Problem Solving Protocol; Sequence Analysis, DNA - methods; DNA methylation; current signals; Nanopore sequencing; attention-based neural network
• ISSN: 1467-5463; 1477-4054; 1477-4054
• DOI: 10.1093/bib/bbae596

Abstract Decoding DNA methylation sites through nanopore sequencing has emerged as a cutting-edge technology in the field of DNA methylation research, as it enables direct sequencing of native DNA molecules without the need for prior enzymatic or chemical treatments. During nanopore sequencing, methylation modifications on DNA bases cause changes in electrical current intensity. Therefore, constructing deep neural network models to decode the electrical signals of nanopore sequencing has become a crucial step in methylation site identification. In this study, we utilized nanopore sequencing data containing diverse DNA methylation types and motif sequence diversity. We proposed a feature encoding method based on current signal clustering and leveraged the powerful attention mechanism in the Transformer framework to construct the PoreFormer model for identifying DNA methylation sites in nanopore sequencing. The model demonstrated excellent performance under conditions of multi-class methylation and motif sequence diversity, offering new insights into related research fields.