The pre-print for Slorado and openfish is out. biorxiv.org/content/10.6...

You can do @nanoporetech.com basecalling with not just NVIDIA GPU, but also using a range of AMD GPU.

Would be useful since popular GPUs are now 💸💸💸💸(& hard to buy),

Work by @bonson-wong.bsky.social

Myloasm, our long-read metagenome assembler, is now published! w/ @mgmarin.bsky.social and @lh3lh3.bsky.social

Very rewarding after > a year of development and countless hours thinking about assembly. Thanks to beta testers, Li lab, and reviewers who gave very helpful feedback.

rdcu.be/famFj

We have released minimod v0.5.0.

Now supports a number of different modifications other than CpG methylation.

🫠

Excited to share our latest preprint: "SWARM: A Single-Molecule Workflow for High-Precision Profiling of RNA Modifications" www.biorxiv.org/content/10.6...
Led by Stefan Prodic #RNA #Epitranscriptomics #Nanopore 1/6

Release slow5curl-v0.3.0 · BonsonW/slow5curl Whats Changed Can now download ex-zd compressed SLOW5/BLOW5 files Update to slow5lib-v1.4.0 Minor usability and documentation improvements Full Changelog: v0.2.1...v0.3.0

Also, @bonson-wong.bsky.social did a slow5curl release github.com/BonsonW/slow... to support these ex-zd compressed BLOW5

Release slow5tools-v1.4.0 · hasindu2008/slow5tools What's Changed slow5tools skim supports the new auxiliary field open_pore_level introduced in latest ONT pod5 slow5tools degrade has new profiles added (by @sashajenner and @hiruna72) and are docu...

slow5tools v1.4.0 released github.com/hasindu2008/...

Many bit profiles for ex-zd lossy compression added by @hiruna72, who reduced 275TB of historical @nanopore rawdata at @GenTechGp to 172TB.

guide to lossy archive: hasindu2008.github.io/slow5tools/a...
paper: doi.org/10.1101/gr.2...

Now that the @nci-australia.bsky.social Gadi supercomputer in Australia has H200 GPUs, @nanopore basecalling is much faster than before. For those who are interested, I have updated my example scripts [https://github.com/hasindu2008/nci-scripts] to demonstrate how these can be used.

congratulations!!!! 🎉

Release slorado-v0.4.0-beta · BonsonW/slorado What's Changed Support for RNA models >= v5.1.0 --flash yes|no option that enables flash attention for transformer models (SUP >= v5.0.0). This is "no" by default to maximise compatibility across ...

slorado v0.4.0-beta is released: github.com/BonsonW/slor...
- Support for RNA basecalling
- 20-35% performance improvement for super-accuracy basecalling

Work by @bonson-wong.bsky.social.

Yes that is the process.

Very detailed step are here:
hasindu2008.github.io/cornetto/doc...

If you run into obstacles when setup, please open an issue - will be helful for us to streamline the setup steps where possible.

Yes, it works with MinION.

I have detailed the steps here: hasindu2008.github.io/cornetto/doc...

Not sure if this is simple enough though :D

I made a small wrapper library for reading slow5/blow5 files in #julialang based on the C slow5lib. If anyone working with #nanopore sequencing is interested, you can find it here: https://codeberg.org/mzdravkov/Slow5.jl

Cornetto v0.2.0 is now released for using programmable selective nanopore sequencing for genome assembly.

GitHub: github.com/hasindu2008/...
Paper: nature.com/articles/s41...
Datasets: hasindu2008.github.io/cornetto/doc...

... and a banner made by Ira Deveson referring to ‘no boring bits’.

Congratulations Hiruna Samarakoon (yet to be on bluesky) for winning the #abacbs2025 “Torsten Seemann” Outstanding Bioinformatics Software Developer Award!!! 🎉🎉🎉

Cornetto uses readfish by @minomatt.bsky.social for adaptive sampling & hifiasm by Hayou, @lh3lh3.bsky.social et al for assembly.

Targeted sequencing and iterative assembly of near-complete genomes - Nature Communications Long-read sequencing enables high-quality genome assemblies, but challenges remain. Here, the authors introduce Cornetto, a method that improves assembly quality, enables genome sequencing from saliva...

Our cornetto work is now published at www.nature.com/articles/s41...

It can do near-T2T assembly using @nanoporetech.com adaptive sampling
- with less 💸
- reference agnostic, so works for non-humans
- not just blood, even saliva

Just presented at #abacbs2025 yesterday.

With Slorado, now you have more choices for GPUs when basecalling @nanoporetech.com sequencing data. This is a work we collaborated on with AMD, led by
PhD candidate @bonson-wong.bsky.social (poster at #abacbs2025) and great to see being highlighted in the AMD blog: www.amd.com/en/blogs/202...

Targeted sequencing and iterative assembly of near-complete genomes - Nature Communications Long-read sequencing enables high-quality genome assemblies, but challenges remain. Here, the authors introduce Cornetto, a method that improves assembly quality, enables genome sequencing from saliva...

Thanks mate

Off the press just couple of hours before the talk
www.nature.com/articles/s41...

😂

The enduring advantages of the SLOW5 file format for raw nanopore sequencing data Abstract. Nanopore sequencing is a widespread and important method in genomics science. The raw electrical current signal data from a typical nanopore sequ

possibly because ONT use their POD5 format which is not great on HDD. They could have adopted our slow5 format which we showed long ago that works for even HDD, before pod5 came in.
Now the pod5 (while indeed better than fast5) is still behind slow5 in this regards
academic.oup.com/gigascience/...

Sometime ago I changed a configuration file so this location is changed to our /data location permanently. Don't know if they have changed this option now.

Thanks mate

Thanks mate.

Yeh, many better ways exist. We just showed how the most naive method can still keep up truck load volumes of data.

In my personal opinion, more popular it is, more care should be taken to avoid compatibility problems. Adhoc implementation-driven changes can be avoided if enough thought is given at the design phase.

Just checked my logs and yes it is true.
The output produced was wrong - A simple program that sums up the signal values.

I just checked release logs of pod5 - 0.0.20 "Fix bug reading data via C API". So it could be a bug that got introduced in before.

@psy-fer.bsky.social @gigascience.bsky.social

Benchmark comparing SLOW5 and POD5 for nanopore raw signal data has now been published at
@GigaScience
academic.oup.com/gigascience/....

Some plots required a log scale - RAM usage and random access time.

For anyone interested, we uploaded another new HG002 @nanopore R10.4.1 (e8.2.1 enzyme) dataset with raw signals and super-accurate basecalls.
Reads lengths: 9.0 kbases median, 17.2 kbases mean

See:
gentechgp.github.io/gtgseq/docs/...