Exciting news regarding the SBWT rust API! We now have an implementation of the subset rank method in this preprint, written Anastasia Diseth. It takes the space down from 5 bits/kmer to 2.5 bits/kmer (double that if both strands are indexed). About 4-5 times slower. www.biorxiv.org/content/10.6...

New Space-Time Tradeoffs for Subset Rank and k-mer Lookup

Given a sequence S of subsets of symbols drawn from an alphabet of size σ , a subset rank query srank( i, c ) asks for the number of subsets before the i th subset that contain the symbol c . It was r...

biorxiv.org

Gravlax: an annotation-independent molecular evidence archive for single-cell RNA-seq https://www.biorxiv.org/content/10.64898/2026.09.18.752708v1

"self-centred men who are argumentative, stubborn, disagreeable and unempathetic are much more likely to own a high-status car such as an Audi, BMW or Mercedes."

Brent Toderian@brenttoderian.bsky.social · 3d ago

“Why do BMW and Audi owners often seem to drive like idiots? Is it the car that makes them behave aggressively behind the wheel, or are specific types of people drawn to such cars as well being more likely to break traffic regulations? Research at the University of Helsinki provides some answers.”

Artificial-intelligence systems can generate hypotheses, design experiments and analyse data — but humans still need to decide what makes sense go.nature.com/3Vn4WUY

AI co-scientists are revolutionizing how research is done

Artificial-intelligence systems can generate hypotheses, design experiments and analyse data — but humans still need to decide what makes sense.

go.nature.com

Explainer posts here: bsky.app/profile/robp...

#SingleCell preprints@prepub-singlecell.bsky.social · 3d ago

Gravlax: an annotation-independent molecular evidence archive for single-cell RNA-seq #SingleCell 🧪🧬🖥️ https://www.biorxiv.org/content/10.64898/2026.09.18.752708v1

It's here! new preprint: Gravlax, an annotation-independent molecular evidence archive for scRNA-seq. A count matrix freezes one annotation. Molecules never change. Instead, gravlax keeps the evidence. Requantify, query, and discover under any annotation. 🧵 www.biorxiv.org/content/10.6...

What excites me most about gravlax is that it breaks the dichotomy between tiny but *very limited* count matrix data that is shared and reanalyzed on a near constant basis and huge and compute intensive raw read/bam artifacts that are often too inconvenient for people to routinely reuse.

Rob Patro@robp.bsky.social · 3d ago

It's here! new preprint: Gravlax, an annotation-independent molecular evidence archive for scRNA-seq. A count matrix freezes one annotation. Molecules never change. Instead, gravlax keeps the evidence. Requantify, query, and discover under any annotation. 🧵 www.biorxiv.org/content/10.6...

My agent achieved to use my printer, ASI confirmed

Jarno N. Alanko@jnalanko.bsky.social · last wk.

I think that the year of the Linux desktop is finally coming. AI is really good at debugging obscure issues that used to require lots of technical expertise to solve. It already fixed an issue with my Bluetooth headphones, and another one with my USB audio card.

The sylph metagenome profiler is v1.0.0! sylph-docs.github.io A new DB format + approach --> huge performance gains: GTDB-R232 (200k species) now takes < 5 GB of RAM and ~30s (2GB fq.gz). Huge thanks to @benjwoodcroft.bsky.social and his ongoing performance efforts (github.com/wwood/weebill)

On est sûr que ça n'a rien à voir avec le (petit ahem) problème de financement de tout ce bazar... C'est comme construire des autoroutes sans avoir les moyens de les entretenir.

Mediapart@mediapart.fr · last wk.

Le patron d’Anthropic appelle à son tour à « temporiser » le développement de l’IA Par @yphilippin.bsky.social

The international chess Federation banned trans women from playing women's chess. CHESS. It's not about fairness. The fairness argument is intended to radicalize people into more violent forms of transphobia.

Chef Kennedy🐸🏳️‍🌈@kennedy.northsky.social · 2w ago

That rock skipping, yes tossing stones across a pond or river, is a gender segregated sport, should tell you something about how gendered sports are not about fairness of physical ability, and is instead about the male ego

“we compared 5 foundation models and two other deep learning models against deliberately simple baselines for predicting transcriptome changes after single or double perturbations. None outperformed the baselines, .. highlights .. importance of critical benchmarking www.nature.com/articles/s41...

Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines - Nature Methods

The analysis presented in this Brief Communication shows that, despite their complexity, current deep learning models do not outperform linear baselines in predicting gene perturbation effects, thus e...

nature.com

Don't want OpenAI to steal your in-progress millennium prize proofs? Invest in homomorphic encryption today!

NVIDIA, our newest Platinum Member, launched CUDA Rust today. You can now write GPU kernels in Rust and compile them straight to PTX. Learn more on their blog: developer.nvidia.com/blog/introdu...

Introducing CUDA Rust: Two Tracks for Writing GPU Kernels | NVIDIA Technical Blog

In September 2026, NVIDIA announced it is leaning into native GPU programming in Rust. CUDA C++ and CUDA Python are mature, enterprise-grade toolchains, and NVIDIA will be growing and maturing CUDA…

developer.nvidia.com

What’s gonna happen if Overleaf decides they’re going to scrape user documents and starts scooping authors on in progress results?

Millennium math problem solved by OpenAI may have "regurgitated" competitors unpublished work... also they spent more than $22 million in computing power in one week to scoop them. techcrunch.com/2026/09/08/o...

OpenAI fought dirty on career-making math problem, says NYU mathematician | TechCrunch

There is a $1 million bounty for the first person providing a solution to the Navier-Stokes existence and smoothness problem.

techcrunch.com

Unfortunately, I don't think there's a need for the "really makes me think" qualifier here.

Tien Chih@tienchihmath.bsky.social · 2w ago

This, in conjunction with the Navier-Stokes debacle, really makes me think these guys are trying to destroy The Academy so that all future research will belong in the hands of these corporate thugs.

We should talk more about what a travesty the bibliography of OpenAI is 16 measly citations for a 165-page paper, of which 5 are just the problem statement. Only 7 are from this century. They failed to cite the method they used. And the guy whom they offered single-authorship & the Millennium Prize

humanities academics watching the tech companies force the mathematicians to realise they were humanities academics all along

Gro-Tsen@gro-tsen.bsky.social · 2w ago

So, sorry, we don't care about problem solving any more: AI bros have turned this into a meaningless trophy hunt. Its worth is now exactly zero. Even if you prove the Riemann hypothesis or P≠NP, I don't give a single f😑ck unless you can explain it in a way that's interesting.

For anyone not following the drama, there's been apparently a big development in ai math, along with some accusations of very bad behavior by OpenAI. But we don't have details yet. If I understand it, Buckmaster and Levent Alpoge made a big discovery on Navier-Stokes. Events go something like this:

Michael Barany@mbarany.com · 2w ago

copying here so it's in-thread too bsky.app/profile/mbar...

Cute! So in general: x%b = (x+c*(x//b))%(b+c) (b=60, c=4)

tef@tef.bsky.social · 2w ago

move over inverse square root, there's a new cowboy in town www.benjoffe.com/fast-time-of...

Here's how to score your next ERC grant: be of the same gender and country as your reviewer 😉🥲 Homophily in grant evaluations www.nature.com/articles/s41...

Homophily in grant evaluations - Nature Human Behaviour

Fuchs-Schündeln and Holub show evidence of a homophily bias in scientific grant application review—evaluators are more favourable to applicants from their own country and, to a lesser extent, of their...

nature.com

First time I've seen a court issue sanctions for prompt injection in a brief. Pro se plaintiff tries prompt injection (here, white text in tiny font) because he thinks the judge will just use some LLM to read the papers and write a ruling. Judge sees through the ruse.

🧬 The VGP Phase I preprint is out! 🎉 The Vertebrate Genomes Project has completed Phase I, delivering 816 of the highest-quality vertebrate reference genomes ever produced. Explore this major new resource for vertebrate genomics: 📄 tinyurl.com/VGPphase1 @vertebrategenomes.bsky.social #Genomics

PanGBank: a large-scale resource of precomputed microbial pangenomes built with PPanGGOLiN https://www.biorxiv.org/content/10.64898/2026.08.05.742796v1

I'm very pleased to be part of this paper on the first-ever complete diploid human genome, led by my (new!) Hopkins colleague @aphillippy.bsky.social and his lab, with contributions from Hayden Ji in my lab plus many other @jhu.edu and @hopkinsengineer.bsky.social faculty and students.

Cell - a Cell Press journal@cp-cell.bsky.social · 2mo ago

New issue alert👉https://cell.com/cell/current On the cover: The Telomere-to-Telomere Consortium has finished the complete, diploid genome of a person. The cover is inspired by a karyotype of the chromosomes in metaphase, representing the two haplotypes as different colors.

Normally I avoid company-associated lectures but this one looks genuinely interesting. Apparently aeonophile is a thing! Never heard of the term before, but I'm sure we've all asked the microbial life span question before. 🦠🧫 www.zymoresearch.com/pages/aeonop...

Aeonophiles: Can an individual microbial cell live for a million years? | ZYMO RESEARCH

Explore how aeonophiles survive for millennia and what they reveal about evolution and biotechnology.

zymoresearch.com

This is some really cool stuff! A chaining dynamic hashtable that uses *less* space than the input (by using quotienting), that actually seems to have quite competitive performance 🤯 And also a faster flat variant has 1.25 cache misses on average (and at most 2 or 3), that has super fast queries.

arXiv cs.DS Data Structures and Algorithms@csds-bot.bsky.social · 2mo ago

Xilin Tang (Cornell University), Yuqi Mai (Cornell University), William Kuszmaul (Carnegie Mellon University), Alex Conway (Cornell Tech): Succinct and Fast Tiny Pointer Hash Tables https://arxiv.org/abs/2607.28892 https://arxiv.org/pdf/2607.28892 https://arxiv.org/html/2607.28892

There's lots of stuff LLMs can't do, but if you want to disbelieve all the recent math, you have to believe a bunch of Fields Medalists with long track records are just straightforwardly lying about the thing they care most about. Here's a recent blog by Gowers: gowers.wordpress.com/2026/05/08/a...

A recent experience with ChatGPT 5.5 Pro

We are all having to keep revising upwards our assessments of the mathematical capabilities of large language models. I have just made a fairly large revision as a result of ChatGPT 5.5 Pro, to whi…

gowers.wordpress.com