People have got to wake up to how bad the situation in US science is right now

My TEDx talk is online youtu.be/XX0pgmV1_YA?... I talk about using detected species to infer undetected species in #eDNA!

One thing I don't understand yet : how come the Venn diagram of "people who are really excited about AI" and "people who have absolutely zero taste" is a circle

Dawkins does strike me as someone who'd call something conscious if it's just sycophantic to him

Why does nobody read Barbara Tuchman anymore? Reading Guns of August and it 'slaps'

We're looking for a postdoctoral fellow to work at the intersection of ecology and data science 🧪 www.tepunahamatatini.ac.nz/posts/postdo... #movingtonewzealand #movingtonz

I came back to the United States, hearing about a very sad news: Craig Venter passed away today. He is a pioneer, successfully sequenced the first Human genome, and tried to create Synthetic Cells. We lost a giant in Science. RIP. @jcvi.org

This blog post is so timely - just ran an internal power study where to get anywhere, I had to assume a d of 3.5...

Reminder: when authors observe Cohen's d = 4, no they didn't. Article now retracted: www.nature.com/articles/s41... Critique: osf.io/preprints/ps... Blog on plausibility of effect sizes: trustworthy.scientific.claims/posts/if-res...

Costco remains my personal hell. They somehow get all of Perth's most meandering, slow walkers, cram them into a hall, and give them the world's widest shopping carts. In-my-way-maxxing

I recently read this great piece on 'software brain', computer people who can only understand the world as a series of databases. There's an extension to this thinking at play here - Meta, Palantir et al. were built by people who didn't think for a second about what they were building

Been trying Gemini and picture mode to see how it suggests re-arranging our rooms full of kids' toys and sports stuff and so far, every suggestion looks like "the same but you threw away half the stuff". Thanks, I guess??

Found a super-chonky seahare today on the beach - look at the size of it!! (Also pictured, a shoe)

21 files changed, 1966652 insertions(+), 525229 deletions(-) I may have been vibe-coding

I've been reading Peter Burke's The Polymath (a history of) and it's making me extra-depressed. People used to be smart. Now every public intellectual is either ragebaiting, uninformed (looking at you, Precht) or straight-up grifting.

We are recruiting a PI in biomedical data science at @bakerresearchau.bsky.social in Melbourne! A fantastic opportunity and do get in touch if any q's! www.seek.com.au/job/91619982...

This reads like something a neonazi 21 year old with a 12 follower substack posts, not the partner of governments (including my native Germany). It's 2026 why am I even surprised

Excited to share our new findings in @science.org on how the DRT3 bacterial defense system uses a reverse transcriptase that builds DNA repeats without a nucleic acid template. Microbes never cease to amaze! www.science.org/doi/10.1126/...

I feel like there shouldn't be a single user (account) associated with AI rewritten code. The original prompt writer will disappear, so others should be able to port software updates into the AI rewrite without barriers. In AI rewrites, no one is responsible by default

Absolute banger: "find that metaSPAdes consistently outperforms MEGAHIT" "Binning refinement, which combines bins from multiple different algorithms, leads to reduced performance" "We further show that CheckM2 systematically overestimates completeness and underestimates contamination"

Naming a product famously based on stolen IP 'rosalind' is certainly a choice. Can you imagine the company culture that thinks this won't be an issue

This would be a fun paper! How much negative and positive selection occurs per day across all kingdoms of life?

native parallel processing with purrr! tidyverse.org/blog/2025/07...

I feel like the longer you stay in academia, the less hireable you get for "other" bioinfo jobs. Industry/NGO professionalises, adds more for-pay tools academics never see (no exp.). Academia instead focuses more and more on writing ever longer grants/papers as funding dwindles. returns are marginal

Just got my credit card bill and with all living costs exploding (AGAIN) life is difficult, difficult, lemon-difficult

GALBA2 walks into the arena. We rewrote our protein-based genome annotation pipeline in Snakemake. Give it a genome + proteins from close relatives → get gene predictions. No RNA-Seq, no GeneMark needed. miniprot → AUGUSTUS, fully containerised, HPC-ready. github.com/Gaius-Augustus/GALBA2

Working for 3 months in for-profit bioinformatics now - what a change in gears! Definitely much more fast-paced environment, with not much tolerance for error. Almost too much daily context switching. Hard problems, but lots of room for solutions and ideas and pushing. This is good

I took the kids to the Beverley Air Show, and among all the loop-di-loops stunt plane flights, they had a helicopter drop a car from the sky. It was as awesome as that sounds like

Metabuli & Metabuli App v1.2 improve novel species classification with higher precision and recall. New light mode is 1.8× faster and requires 50% less storage while keeping precision. New RefSeq, GTDB, HRGM, and HROM databases added. 💾 github.com/steineggerla... 📄 doi.org/10.64898/2026.03.13.711249

Sequence Display generates large-scale protein sequence–activity datasets by simultaneously reading out the sequence of interest and a downstream editing efficiency - perfect for data hungry ML methods! Happy to have played a small part in this!

Good news, everyone! I thought I had a possum living in my roof, which isn't easy to fix: possums are protected & territorial animals, which means they won't move and need help. But i just saw what's living in the roof, and it's just the fattest rat I've ever seen. Yay, that's an invasive species!

🧬 My haplotype paper is out! We show that deviations between within-sample and population-level haplotype frequencies can be used to estimate how many individuals contributed to an eDNA sample. No tissue references needed, just metabarcoding data and some population genetics. #eDNA #PopGen

🚨New preprint and our results are rather concerning.. We find the "boiling frog" equivalent of AI use. Using large-scale RCTs, we provide *casual* evidence that AI assistance reduces persistence and hurts independent performance. And these effects emerge after just 10–15 minutes of AI use! 1/

“Most large companies are spending more time strategizing against their employees than against their competitors.” 🖐️🎤

remarkable how fast things went from 'rewrite everything in Rust!!!' to 'rewrite everything in Rust using Claude!!!'

And another Quarto announcement; I've alluded to it before, but we're making it "official". We've started work on Quarto 2. The blog post has an overview: quarto.org/docs/blog/po... We'll share more in future blog posts, but here's what you can expect from the Quarto 2 dev effort: (1/)

A deep joint-learning proteomics model for diagnosis of six conditions associated with dementia www.nature.com/articles/s41...

me when I clean up the freezer

Yet another article showing how the Australian university sector is falling apart. Pretty much the same story ever since I did in my PhD in 2016. Will anything change? I highly doubt it. "Is Australia’s university empire losing global appeal?" www.abc.net.au/news/2026-04...

Finished reading Murakami's What i Talk About When I Talk About Running and it's so weirdly non-motivational (I just run because it suits me. If it wouldn't suit me, I'd do something else.), so little about trying to impress the reader, I got up and ran the second-fastest 5km since I started running

As the kids on the internet used to say: "this is bait" www.theguardian.com/commentisfre...

You know what's weird? There's a weird kind of job security in academia. Once you have your 3 year contract, you are safe for the duration. I've been permanent contracts since leaving academia, and I've seen dozens, hundreds of people made redundant at the drop of a hat. It will happen to me, too.

Out now in @newphyt.bsky.social with @joannarifkin.bsky.social, @jotlovell.bsky.social, @spicybotrytis.bsky.social & more! Our high-quality pennycress pangenome is a striking example of genome architecture shaping different kinds of genomic variation, including some surprising centromeric movement

New preprint out! Following two international courses in Brazil and Norway, we show that eDNA metabarcoding can be done fully on-site with portable tools, delivering fish and elasmobranch results in just days, even outside a traditional lab www.biorxiv.org/content/10.6... @jonready.bsky.social

After two years in this house it has finally happened: there's a possum living in the roof. spending the night scritching on my ceiling like they walled in the cryptkeeper and forgot about him

Great find at Elizabeth's in Fremantle, but yeah they're the priciest used book shop in Perth

In my new job have co-designed SOPs for people who run Prism etc. - there has to be a final check by us to ensure everything is above-board if it's compliancy-relevant. We'll do it with AI tools, too

I'm both excited and a little terrified. It feels like a wave of AI-assisted tool rewrites is coming to bioinformatics whether we like it or not. We figured that the best we can do is to try to get ahead of it, so we wrote down some principles for doing it responsibly: rewrites.bio

eDNA is possibly the coolest application! Below: read-level placement across a pan-vertebrate mitochondrial pangenome (left) where a clear cluster of read assignments supports the presence of many mammoth reads (right). Processing ~9 billion reads took just 8 minutes!

Alex Kramer, Alan Zhang and friends posted our preprint today. In it, we introduce Panmap, a tool for phylogenetic placement, assembly, lineage abundance estimation, and eDNA assignment using phylogenetic pangenomes. www.biorxiv.org/content/10.6...