Adrien Leger
@adrienleger.bsky.social
220 followers 91 following 9 posts
Director of Modified Bases Research at @nanoporetech.com EMBL-EBI alumni, keen cyclist and father of two. Views are my own. Reposts are not endorsement.
Posts Media Videos Starter Packs
Reposted by Adrien Leger
immagina.bsky.social
Excited to welcome ‪@adrienleger.bsky.social from‪ @nanoporetech.com to our RNA Symposium! 🎤 His talk will focus on how nanopore direct RNA sequencing can reveal multiple RNA modifications at once. Secure your spot at our symposium now 👉 zurl.co/mrhKq
#RNA #ribosome #nanopore
adrienleger.bsky.social
🚨 We’re hiring ! 🚨
Join our cutting-edge research team as a molecular biologist at @nanoporetech.com HQ in Oxford.
Perfect for a fresh PhD or MSc with a couple of years’ experience. Work at the interface of chemistry, molecular/synthetic biology and AI

👉 ejnh.fa.em2.oraclecloud.com/hcmUI/Candid...
Reposted by Adrien Leger
akshatrathi.bsky.social
The world prefers its energy more in the form of electricity than anything else.
Reposted by Adrien Leger
nanoporetech.com
At London Calling, Rosemary Sinclair Dokos and Lakmal Jayasinghe shared how we’re doubling down on what matters: delivering a future-ready platform with proven performance across DNA, RNA, and now proteins. Click here to discover what’s new – and why it matters: nanoporetech.com/news/london-...
Reposted by Adrien Leger
gringene.org
#NanoporeConf Lakmal Jayasinghe

ONT is the *only* sequencing technology that detects RNA modifications directly. They're now up to 8 modifications that can be detected simultaneously.

[This is something I think ONT should be shouting from the rooftops every day]

RNA barcoding is in beta testing.
image showing RNA modification landscape of human ribosomal RNAs for 18S and 28S rRNA; patterns of different modifications are shown throughout the entirety of the sequence. A validation of synthetic constructs demonstrates RNA modification detection accuracies of 95-99%.
Reposted by Adrien Leger
vellamike.bsky.social
New HAC models are 1.4Q more accurate - a 25% error reduction
Reposted by Adrien Leger
vellamike.bsky.social
Big improvements in modification calling accuracy - 6mA increases from Q16 to Q24
Reposted by Adrien Leger
vellamike.bsky.social
Dorado v1.0.0 introduces dramatic improvement to modified base calling speed
Reposted by Adrien Leger
vellamike.bsky.social
Base calling obeys scaling laws in deep learning - a Q28 base caller using standard chemistry is now available on demand.
Reposted by Adrien Leger
nanoporetech.com
Oxford Nanopore sequencing will be ubiquitous for multiomics….and the next step is proteomics.

First panel-based proteomics assays, before driving towards full protein sensing in the future. #nanoporeconf
Reposted by Adrien Leger
stoibs11.bsky.social
More details on new RNA modified bases detectable with @nanoporetech.com sequencing. #NanoporeConf
Reposted by Adrien Leger
stoibs11.bsky.social
Dorado 1.0 🚀release with new RNA modified base models for 2’Ome as well as improvement of all DNA and RNA modified base models including massive improvement in 6mA for open chromatin prediction! Also a lot more non-mods related features😆! #NanoporeConf

github.com/nanoporetech...
Reposted by Adrien Leger
stoibs11.bsky.social
Dorado 1.0 🚀release with new RNA modified base models for 2’Ome as well as improvement of all DNA and RNA modified base models including massive improvement in 6mA for open chromatin prediction! Also a lot more non-mods related features😆! #NanoporeConf

github.com/nanoporetech...
Reposted by Adrien Leger
nanoporetech.com
With significant ML advances, Dorado & MinKNOW now power built-in accurate variant calling, default methylation detection & short fragment mode. Adaptive sampling & T2T assemblies are production-ready — enabling faster, richer, more meaningful insights.
Reposted by Adrien Leger
iiseymour.bsky.social
Dorado is 1.0! 🎉

LC25 release adds v5.2 DNA/RNA models with a new suite of mod models that are higher accuracy, lower FP rates, and significantly faster.

- Introducing dorado variant 👀
- HAC error rate is down over 25%.
- New 2’Ome mod models
- Hopper & Blackwell SUP speed ups

#Nanoporeconf
Reposted by Adrien Leger
nanoporetech.com
Your speakers for London Calling 2025! #NanoporeConf
Reposted by Adrien Leger
stoibs11.bsky.social
Excited for another London Calling! Incredible updates coming over the next two days! #NanoporeConf @nanoporetech.com
Reposted by Adrien Leger
ewanbirney.bsky.social
I expect Rosemary and Lakmal (the new R&D team leads) will reveal more than just all this - I am looking forward to the reports on Wednesday and looking forward to catching up with everyone nanopore on Friday!
Reposted by Adrien Leger
ewanbirney.bsky.social
Nanopore now can call 4 RNA modifications - m6A (I love how DNA and RNA have different writing conventions - molecular biologists, just don't go changing!) - m5C, pseudourodine, and inosine. This is all new stuff, and there quite a bit on truth sets, sensitivity/specificity to play out
Reposted by Adrien Leger
ewanbirney.bsky.social
The final, most exotic stuff is RNA - so called "direct RNA" where modifications of the RNA is the really exciting stuff (the sequence itself, including intron/exon structure is probably still done best via cDNA, "old school").
Reposted by Adrien Leger
ewanbirney.bsky.social
Basically molecular biologists can either use incorporation (eg, BrdU) or modification (eg, bacterial enzymes tethered to antibodies) or chemical modification (eg, radicals etc) and then the read out is the DNA (or RNA, see below).
Reposted by Adrien Leger
ewanbirney.bsky.social
Nanopore can read other mods as well - hydroxymethylation, the oxidation product of methylation which is only really present in neuronal tissue in vertebrates (watch this space for insights on this - we will be preprinting soon!) and bacterial DNA methylation, like 6mA.