Marcus Stoiber, a Principle Algorithms Researcher with Oxford Nanopore Technologies, presented at the Nanopore Community Meeting 2022 the “Nanopore modified base update.” Stoiber explained Oxford Nanopore sequencing allows for powerful epigenetics research because it can “directly detect methylation with high reproducibility and low bias.” You can sequence native DNA and RNA to assess modification. The algorithms they developed are named Remora, an epi-basecaller. The Remora upgrade picks up signal around motifs of interest and uses a neural network algorithm to evaluate whether it is modified. Stoiber mentioned that there are native modified bases, modified, and enzymatic modified base data types. The production model has improved accuracy per-read, per-site. Stoiber mentioned that bisulfite modification either happens or doesn’t and Remora provides more information and you can set thresholds. The 5-methyl-C all-context model is ready and others are in production. The Remora upgrade can be installed with pip install — upgrade remora and Remora 2.0 is now available. It has a unified interface with Dorado. Now Remora can be run on duplex modified bases on both strands. I now want to try Remora on some bacterial and metagenomic datasets to learn more about it and what it does with these samples. These tools may be really interesting to use with Delftia phage…
