Tonight, I watched the Nanopore Learning Knowledge Exchange about Direct RNA Sequencing. Libby Snell and Botond Sipos explained how the kit and process work. Snell began by addressing the question: why use direct RNA sequencing. They explained that it is truly native RNA sequencing since the RNA strand is read directly in the Nanopore. There is an optional reverse transcriptase step. The advantages of direct RNA sequencing are several: information is not lost because poly(A) tail and modified bases are detected, and the kit is relatively straightforward. Snell explained that the direct RNA sequencing kit uses an RNA-specific motor. Direct RNA sequencing can be used for transcriptomics, differential gene expression, differential transcript usage, genome annotation modifications, poly(A) tail length estimation, structural variations, and sequencing viral genomes (as single reads!). The SQK-RNA002 is an upgrade of the SQK-RNA001, released in 2017. Snell did recommend doing the reverse transcriptase step. The kit was optimized using Thermo Fisher reverse transcriptases. The cDNA is not sequenced because it is pealed off above the pore. It is possible to do the kit without the cDNA step. This reduces time but also throughput. Snell recommended targeting transcripts >200 nt in length, and there is no upper limit. Prokaryotic transcripts can be polyadenylated. Sequence-specific transcripts can be used. The kit can be stored at -20 instead of -80 C and has double the output of the initial kit. The kit has six preps and requires 500 ng of poly(A) RNA per sample. Snell recommends a RIN above 7 for RNA. ONT is recommending RNA extraction protocols. They do recommend enriching kits that enrich poly(A) transcripts. Using RNase-free environments and reagents is essential. Multiplexing is being developed, as well as VolTRAX kits for RNA sequencing. Sipos spoke about using “ERCC datasets” with known transcript amounts to gain confidence in methods. They recommend using the Scotty web tool for designing RNA-Seq experiments. As a pilot experiment, Sipos recommended two replicates per group and 10 million reads (2 PromethION chips for dRNA). Scotty helps users learn about the required depth and sample size. Sipos also recommended a pipeline that is available on the Nanopore GitHub page. The tools include input transcriptome and references, alignment and counting, differential expression analysis, and output tools. Salmon is used for quantification. Sipos explained that an input genome in fasta and direct RNA sequencing can be used for annotation and visualization. They also have a pipeline for annotation that uses a genome in fasta format and reads in fastq. The pipeline is called Pinfish and has polishing and collapsing of transcripts. Sipos noted that the FLAIR tool by Workman and colleagues does genome annotation using direct RNA data. Poly(A) tail length estimation can be performed with Nanopolish. A tool to detect modifications in native RNA is Tombo, including 5mC and by comparison of sample and control. I know RNA004 is in early access and also noticed RNA flow cells for both the MinION and PromethION. I hope we can continue to try direct RNA sequencing with bacterial samples!
