Billy Lau from Stanford University School of Medicine spoke at the Nanopore Community Meeting 2022. Lau’s 20-minute session was entitled “Single molecule methylation profiles of cell-free DNA in cancer with nanopore sequencing.” The study had the goal of exploring the “potential utility of nanopore sequencing for assessing cancer in cfDNA.” Lau explained that the conventional methods for cancer screening require invasive colonoscopy and monitoring with CT imaging. However, screening has poor adherence and is procedurally intensive. Lau explained that liquid biopsies have been explored as an option for non-invasive cancer screening. Epigenetic reprogramming in cancer can be detected by analyzing modifications. With traditional Illumina methylation analysis, bisulfite treatment is used but it causes DNA damage, PCR amplification and GC skew, and complicated deduplication. Lau spoke about the benefits of long-read sequencing to directly detect modifications and also mentioned challenges with sample input requirements. Lau and colleagues designed a high-throughput sequencing workflow. They used the kit 9 chemistry and 96 native barcoding kit and sequenced on PromethION flow cells, generating 10 Tb of data. Lau mentioned that it would take about 40 days to press had they not used 25 nodes on the Stanford HPC. They used Guppy v5 hac and Megalodon 2.4.0. They conducted a trial with 20 patients and 1 ml plasma samples. Lau and team noticed methylation signatures in cfDNA that were found in cancer patients. The team next wanted to know if they could gain insight into tumor burden by sequencing cfDNA. For this, they used a classification system for single molecule reads. I am impressed tha tthey were able to “confidently” classify reads as tumor specific. Lau concluded that they have a cfDNA methylation analysis system and will now apply this to larger cohort studies and for biomarker discovery. The techniques used here were impressive and could be applied to other settings and systems, including microbial and phage base modifications.
