Happy to share that RISER is published! RISER enables real-time biochemical-free enrichment or depletion of RNA classes during direct RNA sequencing.

RISER has been updated to support @nanopore RNA004 adaptive sampling to enrich or deplete RNA classes such as mRNA, mtRNA and globin mRNA:

@nickschurch Re truncated isoforms - it depends how short they are. If they are too short then they will be fully sequenced before RISER gets a chance to make a decision.

@nickschurch Thanks for the question. We trained a CNN to predict, from the start of the nanopore signal (2-4 secs), whether the RNA is from the target class or not. No basecalling needed. 2-4s corresponds to ~140-280nt (R9.4.1 pore). We currently have models for mRNA, mtRNA and globin mRNA.

@Psy_Fer_ @martinalexsmith @nanopore Yeah that sounds good! Feel free to shoot me a message or email when you’d like to catch up :)

@Psy_Fer_ @martinalexsmith @nanopore Interesting! For me when it happens it is always from the start of the run - and restarting the ReadUntil client and/or the MinKNOW GUI doesn't resolve it. Is it happening for you mid-way through a run?

@Psy_Fer_ @martinalexsmith @nanopore I also tested changing break_reads_after_seconds right at the start of the project, but then was told by some of the adaptive sampling team at ONT that changes to this param can impact/break other parts of MinKNOW - so best to leave as is :)

@Psy_Fer_ @martinalexsmith Yeah that's happened to me before too. I haven't found the root cause yet. The quickest solution has been to change MinIONs - but was planning to contact @nanopore about it, since it appears to be a MinKNOW/ReadUntil issue. Please let me know if you solve it!

@Psy_Fer_ @martinalexsmith Thanks for asking, yes! Retraining is currently underway! More to come soon :)

@Psy_Fer_ @martinalexsmith Thanks James !

@danielle_gutman Thanks Danielle!!

@shaam_alabed Thanks Shaam ! :)

@shibi76 We don't yet have a model for tRNA - tRNA is difficult to sequence with nanopore due to its short read lengths (although there is Nano-tRNAseq from the @NovoaLab). It may be possible to use in conjunction with RISER, but we haven't explored this. I hope this helps!

RT @sgqcheng: .@EduEyras @alex__sneddon develop RISER, a biochemical-free technology to enrich or deplete RNA classes in real-time during d…

@EduEyras @JCSMR @CCBS_ANU @ANUbiomolecular RISER’s software is modular and retrainable, to enable easy adaptation to other RNA classes or sequencing chemistries. RISER is available on GitHub: