Chen Yong from Oxford Nanopore helped facilitate the Knowledge Exchange session I watched tonight: “Decoding the epigenome with Oxford Nanopore real-time methylation detection.” This webinar was recorded. Sayonika Mohanta from ONT explained that methylation is central to gene expression. They spoke about the role of methylation in gene expression: her or hypo methylation can affect levels of expression. 5mC modification can occur in CpG islands. Mohanta summarized some of the known impacts of 5mC modification. While bisulfite conversion and sequencing require several steps and high input, Nanopore offers native DNA/RNA methylation detection with “high reproducibility and low bias.” Nanopore has Reduced Representation Methylation Sequencing (RRMS) that takes advantage of adaptive sampling. Arthur Rand, a Research Scientist in Machine Learning at Oxford Nanopore Technologies, explained how chemical modifications are sensed and detected by the nanopore. Rand explained that with MinKNOW software, you can select the detection of modified bases. Rand shared the command line code to use dorado basecaller and create a bam file with basecalls. Rand noted that they use synthetic molecules to sequence and create models for modified bases. The team has developed a command-line tool called modkit that can help analyze modified bases. One option is to create a mod.bam file. The advantage of modkit is that it was designed to filter and explore data. The modkit can provide filtered or ranked modifications at sites or in regions. Hemi-methylation is also possible. Rocio Esteban is an Application Support Bioinformatician with Oxford Nanopore Technologies. Esteban spoke about the applications of methylation analysis and which modifications can be detected currently. One can perform an analysis to identify sites when characterizing genomic imprinting using the EPI2ME human variant workflow. Esteban also shared several applications of this approach in cancer research. Esteban shared an example of using Nanopore age predictions. They noted that the accuracy is around 99%. Yong shared that there is additional information on the ONT website.
