Tonight I watched the Nanopore Community Meeting Houston session by Bryant Catano, ONT Product Specialist. The session was a Masterclass entitled “How to get started with data analysis.” Catano spoke about the variety of options for data analysis available, from point-and-click to command line. Tools are available that are accessible, scalable, and versatile. Catano noted that you can run your analyses on the cloud, on a server, on a cluster, or locally. Basecalling uses neural network models to convert raw signal to sequence reads. Improvements are disseminated through software updates. The standard FASTQ file is used for sequences and is amenable to compression. POD5 files are used for the raw data and can be used for re-basecalling. The files are considerably larger than FASTQ files. BAM files are used for alignments, for example. POD5 is an optimized file format that has faster read/write performance. FASTQ files start with a label line and metadata. Catano explained that typical bioinformatics workflows are all about file format conversions. Catano shared a table with different ONT tools including EPI2ME and MinKNOW, both with graphical user experiences. MinKNOW basecalling has fast basecalling, high accuracy, and super accuracy modes. Methylation detection can be enabled in MinKNOW when beginning a run. MinKNOW provides real-time run metrics. EPI2ME in the cloud or locally run are options. There are also tools on GitHub. PCR-free native reads can be used to identify methylation. Alignments can be visualized with popular genome browsers. Methylation frequency can be observed between haplotypes: chromosomal portions coming from one of the two parents. Targeted methylation analysis is possible with adaptive sampling. This is “enrichment in silico” and can be very powerful, but also computationally intensive. Catano spoke about Reduced Representation Methylation Sequencing (RRMS) to enrich for target areas, sequence, and detect methylation. RRMS could be an interesting and exciting lab to do with students and unique samples. Catano concluded that “no bioinformatic expertise is needed for your needs.” There are several user-friendly tools that offer powerful analyses.
