New preprint from the lab! "Planetary structure and drivers of diazotroph communities reveal key reservoirs of nitrogen-fixation potential" Nitrogen fixation is scattered right across the prokaryotic tree — but it's heterotrophs, not cyanobacteria, that dominate the potential. 🧵

bioRxiv Microbiology@biorxiv-microbiol.bsky.social · last wk.

Planetary structure and drivers of diazotroph communities reveal key reservoirs of nitrogen-fixation potential https://www.biorxiv.org/content/10.64898/2026.07.27.741065v1

Looking for BGCs in large metagenomic datasets? Our new biorxiv preprint introduces metaSMASH, a scalable fork of antiSMASH designed specifically for metagenome-scale BGC detection and analysis : www.biorxiv.org/cgi/content/... Thanks @canerbagci.bsky.social and @kblin.bsky.social ❤️

metaSMASH: Scalable Biosynthetic Gene Cluster Detection for Large Metagenomic Assemblies

antiSMASH is widely used for biosynthetic gene cluster (BGC) detection and annotation, but its standard workflow is poorly suited to large metagenomic assemblies, where massive contig counts create severe runtime bottlenecks and complicate downstream result exploration. We present metaSMASH, a re-engineered fork of antiSMASH for metagenome-scale BGC analysis. metaSMASH preserves the original antiSMASH detection and annotation logic while introducing streaming, memory-bounded execution, record-level parallelisation, optional output filtering, and an interactive dashboard for large result sets. Across 25 benchmark metagenome datasets, metaSMASH reproduced identical BGC detection results while dramatically reducing computational cost. Relative to the default antiSMASH configuration, metaSMASH was a geometric-mean 38x faster. It also outperformed an ad hoc chunked antiSMASH workflow: in the default configuration it achieved a geometric-mean 2.9x speed-up and 1.7x lower peak memory, and with extended-analysis modules enabled it was 2.7x faster and used 3.1x less memory while completing all datasets, whereas the ad hoc workflow ran out of memory on the two largest assemblies. By substantially reducing the computational burden of large-scale metagenome analysis without sacrificing result equivalence, metaSMASH makes routine mining of assembled metagenomes more practical and provides a scalable foundation for natural product discovery from complex microbial communities. ### Competing Interest Statement The authors have declared no competing interest. German Center for Infection Research, TTU Novel Antibiotics 09.716 Volkswagen Foundation, 0072511-00

biorxiv.org

New lab preprint - Deep learning models are most often difficult to interpret black box models. What if we engineer them with interpretation in mind? Dennis Gankin carefully studied so-called biologically inspired neural networks and found an interesting phenomenon www.biorxiv.org/content/10.6...

Leveraging multiplicity in biologically informed neural networks to uncover disease heterogeneity

Biologically inspired neural networks (BINNs) embed pathway, ontology, or protein-interaction structure directly into neural networks, promising interpretable disease prediction where hidden nodes map...

biorxiv.org

I wish I could teach people just how grating it is to get the fourth clearly AI job application of the week that reads like "While I'm interested in X, to be honest what truly drew me to your lab was Y. You not only do A; you also do B."

Great to see this important study published. For some time I've pointed to the preprint version when students ask about co-evolution between humans and their microbiota. "Limited codiversification of the gut microbiota within humans"

I've been thinking of how to make tools "AI-ready" so it's not just an empty buzzword I'd like to hear other opinions, but this is what I'm trying with SemiBin: 1) bundle skills with the tool SemiBin2 install-skills will (in the next release) install a skill for claude/codex/...

I remember reading the snarky "Reading the entrails of chickens: molecular timescales of evolution and the illusion of precision" during my PhD 20 years ago. nsmn1.uh.edu/dgraur/Artic... pubmed.ncbi.nlm.nih.gov/14746989/ certainly a classic takedown, but it's great that the methods advanced!

Reading the entrails of chickens: molecular timescales of evolution and the illusion of precision - PubMed

For almost a decade now, a team of molecular evolutionists has produced a plethora of seemingly precise molecular clock estimates for divergence events ranging from the speciation of cats and dogs to ...

pubmed.ncbi.nlm.nih.gov

ssolo.bsky.social@ssolo.bsky.social · last mo.

Ancestral genome reconstructions get noisier the deeper in time you go. The usual response: distrust them and joke about reading entrails. Ours is to train on the noise! The result is calibrated phenotype prediction back to the LBCA deep in the Archaean. New preprint www.biorxiv.org/content/10.6...

We've discussed this guide in the lab: zenodo.org/records/1845... and found it a good starting point to responsibile AI usage. Researchers still need to understand all of their pipelines and be answerable for their output, which is impossible if all of it is AI and agent generated

From their examples, it unfortunately looks more like an abdication of domain expertise to AI rather than just handling friction points. Like the three sentence prompt to write a review article 🤢 and being proud of a host of AI agents having reviewed the papers rather than oneself

Claude Science - an AI workbench for science. This looks like an interesting attempt at making scientific workflows easier and more reproducible. I'm curious to give it a try. www.anthropic.com/news/claude-...

Claude Science, an AI workbench for scientists

Claude Science is a customizable app that integrates the tools and packages researchers most often use, produces auditable artifacts, and provides flexible access to computing resources.

anthropic.com

Don‘t forget to register for what promises to be an end-of-the-year highlight in functional microbiome research 👇🏻 !

Typas Lab@typaslab.bsky.social · last mo.

How can we harness the vast functional diversity of microbes on our planet? Join us at EMBL Heidelberg from 8–10 December 2026 to hear about ongoing efforts to decode gene function and organisation in human gut microbes, and to present your own research. www.embl.org/about/info/c...

Universal microbiome signature for CRC Large gut microbiome meta-analysis of colorectal cancer uncovers CRC microbiome signatures universal across ages & sequencing methods & modifiable by dietary interventions, ie fiber. www.cell.com/cell-host-mi...

Meta-analysis reveals microbiome signatures for colorectal cancer that are universal across age groups and sequencing methods

Pekel, Karcher, et al. present the largest gut microbiome meta-analysis of colorectal cancer to date, integrating 6,779 amplicon and shotgun sequencing profiles. They establish a robust microbiome sig...

cell.com

People always say "tell a story" but then don't explain how to do so! How is that helping? If we knew how to tell a story we wouldn't need the advice in the first place. 1/

Stephen Turner@stephenturner.us · last mo.

How to give a bad talk www.nature.com/articles/s41...

More evidence, from a large-scale study in China, that using AI hurts learning if it undermines mental effort. When homework time drops due to AI use, so do test scores. Across studies, there is a clear theme: AI tutoring in support of classes is good, using AI to "help" with homework is bad.

When utilized in literature review, LLMs consistently 1. fail to mention female authors in female-led literatures, 2. insist that men are more influential or more heavily cited when this is contradicted by objective citation counts, and 3. attribute women’s work to hallucinated male scholars.

they even got the gradients right by pouring the H2O2 in from the side. it's a giga-plate fr @baym.lol is seeing his vision fully realized

Harmit Singh Malik@harmitmalik.bsky.social · 2mo ago

One of the world's most tightly controlled experimental evolution projects is underway in DC, and I hope someone has had the foresight to collect a time series of algal samples. www.cnn.com/2026/06/16/p...

I'm attending a digital humanities event in Montreal. There was a keynote on AI and something about the talk made me wonder if it had been written by Claude. I said as much in the Q&A. As I posed the question, the speaker shifted, looking slightly uncomfortable. What he said next shocked the room +

Interesting and thoughtful thread about thé poisoning of academic discourse (my interpretation) by AI

Crystal Fleming クリスタル@alwaystheself.bsky.social · 2mo ago

I'm attending a digital humanities event in Montreal. There was a keynote on AI and something about the talk made me wonder if it had been written by Claude. I said as much in the Q&A. As I posed the question, the speaker shifted, looking slightly uncomfortable. What he said next shocked the room +

😎 to have this out! --> stay tuned, this is a series about 📊#dataviz for #science 🧪

Nature Methods@natmethods.nature.com · 2mo ago

We are very excited to bring back our popular Points of View column, written by @helenajambor.bsky.social! Check out her first column in our July issue about the usage of color scales in scientific figures. www.nature.com/articles/s41...

I’ve officially resigned as Associate Editor for Frontiers in Systems Neuroscience. It used to be a reputable journal, but became a case study in how forced automation destroys academic integrity. 👇

Our new paper „Variations in the latitudinal diversity gradients of the ocean microbiome“ was just published in @cp-cellhostmicrobe.bsky.social 🥳 We present a global analysis of prokaryotic #biodiversity in the surface and mesopelagic layers of the ocean 🦠🌊 www.cell.com/cell-host-mi...

Variations in the latitudinal diversity gradients of the ocean microbiome

Eriksson et al. conduct a global marine microbiome diversity analysis and demonstrate how the surface ocean latitudinal diversity gradient is disproportionately shaped by a few species-rich taxa. Mask...

cell.com

My first EMBL project uncovering synthetic steroid bacterial metabolism is now available publicly! My favourite part is the discovery of novel desmolase enzyme - see the preprint how we have found it 🧬💊🦠

bioRxiv Microbiology@biorxiv-microbiol.bsky.social · 2mo ago

Bacterial metabolism of synthetic steroids across ecosystems reveals diverse biotransformation products, reactions, and enzymes https://www.biorxiv.org/content/10.64898/2026.06.05.730323v1

I wrote about AI in academia. "PhD-level thinking", LLM bias, grunt work, alignment, AGI, data center water use, AI politics -- something for everyone.

Thoughts on AI in academia

PhD-level thinking, LLM bias, alignment, AGI, data centers, and AI politics

open.substack.com

1. How common is LLM use in scientific publishing, and how does it vary across field, publisher, journal prestige, author demographics etc.? @kylesiler.bsky.social has new paper in PNAS that addresses this question on a massive scale: 7.3 million papers from Elsevier, PLOS, MDPI, and Frontiers.

The diffusion of large language models in published academic articles | PNAS

Large language models (LLMs) are rapidly changing academic research, raising questions of who is adopting these tools and under what conditions. Th...

pnas.org

NEW: Hackers say that they used Meta’s AI support chatbot to break into a host of high-profile Instagram profiles by asking the support bot to change the email address associated with the target account.

Hackers Simply Asked Meta AI to Give Them Access to High-Profile Instagram Accounts. It Worked

The exploit shows the extreme risk of offloading technical support to AI.

404media.co

New in @natgenet.nature.com: a moving obituary for Peer. Francesca Ciccarelli, Jan Korbel & Christian von Mering reflect on his scientific vision, from protein domains and genome evolution to metagenomics, and on his generosity as a mentor and friend to so many. www.nature.com/articles/s41...

Peer Bork (1963–2026) - Nature Genetics

Nature Genetics - Peer Bork (1963–2026)

nature.com

Our work exploring host-microbial co-metabolism of bile acids in patients with alcohol-related liver disease (ALD) has been published in the JHEP Reports: www.sciencedirect.com/science/arti... 🪐 A big thank you to everyone involved, to the GALAXY and MicrobLiver consortia who made this possible! 🙌

Alcohol-related liver disease disrupts bile acid homeostasis and gut microbial bile acid metabolism

Alcohol overuse disrupts liver function and alters gut microbial communities, with alcohol-related liver disease (ALD) causing half of all liver-relat…

sciencedirect.com

And we are live! Massive thanks to everyone who contributed to this. Reading your posts made me so much madder than I ever knew was possible

The 40 Most Rage-Inducing Problems in Tech

The bugs, broken apps, and nightmare customer-service bots we can't escape, presented as a blessed and sacred addendum to Pope Leo XIV’s new encyclical on AI

theringer.com

Brian Phillips@brianphillips.lol · 2mo ago

I'm working on a column about the tech annoyances that drive us crazy, and I want it to be as universal as possible, so tell me yours! E.g. scanning a QR code to read a menu, never receiving the one-time passcode they supposedly texted you, "verify you're human" by IDing tiny motorcycles, etc.

Our study describing `microntology` and annotations for >300k metagenomes is now online in Bioinformatics @academic.oup.com : academic.oup.com/bioinformati... Congratulations to @fullam.bsky.social @vishnuprasoodanan.bsky.social & @biocs.bsky.social ! See the thread below for more details.

microntology: a lightweight, data-driven controlled vocabulary to describe Earth’s microbial habitats

AbstractMotivation. Data-enabled studies of microbial ecology and evolution depend on high-quality descriptions of microbial habitats, based on curated and

academic.oup.com

Sebastian Schmidt@tsbschm.bsky.social · 7mo ago

We just released a @biorxivpreprint.bsky.social describing the *microntology*, a lightweight controlled vocabulary that we have been using for a while to tag metagenomes with contextual data. Study by @fullam.bsky.social @vishnuprasoodanan.bsky.social et al: 1/ www.biorxiv.org/content/10.6...

A very important preprint to check out: The holobiont is not a useful model for most host-microbiome interactions By Gavin M Douglas an S. Andrew Inkpen ecoevorxiv.org/repository/v...