Happy to share our new work on Temperature-driven biogeography of marine viruses infecting Micromonas:
academic.oup.com/ismecommun/a...
@gretelwaugh.bsky.social
Microbial generalist π¦ | Keen freediver & birdwatcher | Reluctant #Rstats enthusiast | she/her
Happy to share our new work on Temperature-driven biogeography of marine viruses infecting Micromonas:
academic.oup.com/ismecommun/a...
Excited to annouce the latest research published by our lab, which is led by Dinesh Subedi and is out today in Nature Microbiology.
www.nature.com/articles/s41...
"It would be a foolish enterprise, and doomed to failure.."
18.09.2025 14:43 β π 2 π 0 π¬ 0 π 0An excerpt from a scientific manuscript, Monod, 1949, with the following phrase highlighted in yellow: 'No attempt is made at reviewing the literature on a subject which, as we have just seen, is not really a subject at all.'
Oh, to have been a scientist in the 1940s...
(Monod, 1949)
We present our new preprint titled "Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation". We quantify LLM hacking risk through systematic replication of 37 diverse computational social science annotation tasks. For these tasks, we use a combined set of 2,361 realistic hypotheses that researchers might test using these annotations. Then, we collect 13 million LLM annotations across plausible LLM configurations. These annotations feed into 1.4 million regressions testing the hypotheses. For a hypothesis with no true effect (ground truth $p > 0.05$), different LLM configurations yield conflicting conclusions. Checkmarks indicate correct statistical conclusions matching ground truth; crosses indicate LLM hacking -- incorrect conclusions due to annotation errors. Across all experiments, LLM hacking occurs in 31-50\% of cases even with highly capable models. Since minor configuration changes can flip scientific conclusions, from correct to incorrect, LLM hacking can be exploited to present anything as statistically significant.
π¨ New paper alert π¨ Using LLMs as data annotators, you can produce any scientific result you want. We call this **LLM Hacking**.
Paper: arxiv.org/pdf/2509.08825
"We are told that AI is inevitable, that we must adapt or be left behind. But universities are not tech companies. Our role is to foster critical thinking, not to follow industry trends uncritically." www.ru.nl/en/research/...
12.09.2025 10:45 β π 874 π 374 π¬ 11 π 20How a Group of Students in the Pacific Islands Reshaped Global Climate Law www.nytimes.com/2025/09/10/m...
11.09.2025 22:33 β π 5 π 3 π¬ 0 π 0This is figure 3, which shows projected changes in annual phytoplankton biomass production between pre-industrial (1860) and future (2090β2100) conditions, based on alterations in temperature, nutrient supply, mixing and ocean circulation anticipated under a moderate and high warming scenario.
Prochlorococcus β Earthβs smallest and most abundant photosynthetic organism β may see population reductions of up to 51% in tropical oceans by 2100 under moderate and high warming scenarios, according to a study in Nature Microbiology. go.nature.com/3V3mvGn π π§ͺ
08.09.2025 19:01 β π 42 π 24 π¬ 1 π 3Happy phage day! π¦ Special issue on ecology and evolution of bacterial immune systems, including this paper from @josie-e.bsky.social et al exploring CRISPR - phage interactions. π§΅β¬οΈ
05.09.2025 10:14 β π 3 π 0 π¬ 0 π 0π€―
03.09.2025 10:07 β π 3 π 0 π¬ 0 π 0π¨I have a fully funded 4-year #PhD opportunity in my lab on 'Tuneable multi-input genetic logic circuits for designer probiotics'.π¦ We'll be throwing all the tech we have at this one to make reliable engineering of diverse microbes a reality. Applicants must be UK national...
03.09.2025 09:00 β π 3 π 4 π¬ 2 π 0PhD opportunity, please share:
We are looking for a prospective PhD student to start in October 2026 who is excited about bacteria (Klebsiella pneumoniae), how they interact, and how they exchange DNA.
All details can be found on the funder website π gw4biomed.ac.uk/developing-c...
Exciting new metabolism unlocked - sulfur oxidation coupled to iron reduction! Big congrats and very happy for Songcan, Marc, Alex and team! :-)
28.08.2025 11:12 β π 3 π 1 π¬ 1 π 0Very happy to see this one out: doi.org/10.1093/molbev/msaf192
Here, we assessed the contribution of natural transformation to the acquisition of novel genes. See preprint thread. @molbioevol.bsky.social #microsky #evobio
New paper out @molbioevol.bsky.social on antimutator Ecoli, mutational spectra and the distribution of fitness effects of antibiotic resistance with @knightjar.bsky.social @rokkrasovec.bsky.social and others @mermanchester.bsky.social doi.org/pzfm
Readπ§΅for highlights and fun stats (1/13)
Out yesterday. Using a translation reporter on a mini-tn5 transposon, and high freq. mutagenesis, we looked at translation on whole genome scale independently of genome annotation. academic.oup.com/nar/article/...
14.08.2025 19:00 β π 33 π 16 π¬ 4 π 1Horton, Cherry, @gretelwaugh.bsky.social @taylorlabgroup.bsky.social et al. combine experimental evolution and bioinformatics in bacteria to characterize a β₯8β
bp motif that can increase T:AβG:C mutation rates >1000-fold than the baseline rate.
π doi.org/10.1093/molbev/msaf183
#evobio #molbio
Our new paper exploring mutational hotspots in bacteria π§¬β‘. I learnt lots as part of this project; unpacked in James' π§΅(below). Very excited to see it finally in the wild!
14.08.2025 13:10 β π 3 π 0 π¬ 0 π 0New paper out! In it i review the context and implications of the "chitin raft hypothesis" for how cyanobacteria colonized the open ocean.
Included is a new perspective on why there were two major stages of atmospheric oxygenation over Earth history.
royalsocietypublishing.org/doi/10.1098/...
Great Barrier Reef suffers biggest annual drop in live coral since 1980s after devastating coral bleaching
05.08.2025 15:01 β π 50 π 44 π¬ 6 π 8New paper that may be of interest to ecologists studying global change, disturbances, biodiversity, ecosystem functioning, or response diversity: doi.org/10.1038/s414...
02.12.2024 07:45 β π 84 π 29 π¬ 2 π 1Very proud to have our labβs first work published in @cp-cellhostmicrobe.bsky.social!
In this work, we use functional metagenomics to find phage defenses from human and soil microbiomes!
Congrats to first author @luis840alberto.bsky.social !
www.cell.com/cell-host-mi...
(1/3)
New pre-print from my lab!
Plasmids link antibiotic resistance and phage defense.
E.coli plasmids are hotspots for both antibiotic resistance and phage defense. Phage therapy has the potential to accidentally select for antibiotic resistance!!!
#microsky#AMR#phage
www.biorxiv.org/content/10.1....
#newpaper now out in @nature.com!
A huge expansion of the human niche in Africa ~70 kya likely equipped later #outofAfrica dispersals with a unique ecological flexibility.
www.nature.com/articles/s41...
led by Emily Hallett, @mikleonardi.bsky.social, Andrea Manica, @elliescerri.bsky.social 1/4
Our article showing energy production from air alone is in PNAS today. This paper provides an ultimate biochemical proof that microbes can survive simply by 'making' energy from air (making two ATP per molecule atmospheric H2 consumed) and has biotech applications. www.pnas.org/doi/10.1073/...
25.07.2025 00:54 β π 118 π 55 π¬ 4 π 6Looks really cool, thanks Tiff!
24.07.2025 09:51 β π 0 π 0 π¬ 0 π 0graphical abstract of the article the extended mobility of plasmids
Here's our new broad review on the extended mobility of plasmids, about all mechanisms driving and limiting their transfer. From conjugation to conduction, phage-plasmids to hitchers, molecular to evolutionary dynamics, ecology to biotech. The state of affairs. 1/9 academic.oup.com/nar/article/...
23.07.2025 07:35 β π 185 π 94 π¬ 4 π 9Conference flyer, all info is on link too
Please share! π¦
Workshop on Phage Therapy in Liverpool on 25th November organised by Stineke van Houte, @jojofoth.bsky.social, @brockhurstlab.bsky.social, & Edze Westra.
Limited FREE tickets & more info here: sites.exeter.ac.uk/vanhoutelab/...
#MicroSky
π§΅ Do you have R packages that make your workflow smoother or help you avoid problems you used to face all the time? Iβve collected a few that solved issues that used to trip me up π
18.07.2025 19:09 β π 18 π 10 π¬ 1 π 1Logo for the Sandpiper website
Out in @natbiotech.nature.com: Metagenome taxonomy profilers usually ignore unknown species. SingleM is an accurate profiler which doesn't, even detecting phyla with no MAGs. Profiles of 700,000 metagenomes at sandpiper.qut.edu.au. A π§΅
16.07.2025 21:59 β π 129 π 71 π¬ 7 π 9