Aaron Weimann

@aweimann.bsky.social

Researcher at Uni Cambridge working on bacterial evolution and AI

I’m at ECCB 2026 this week! I’ll be presenting my new eukaryote pangenome simulator, PansimNuc (github.com/samhorsfield...), at Poster C-G.38 on Wednesday 2nd September - drop by if you’re interested in modelling pangenome evolution! @eccb-europe.bsky.social

GitHub - samhorsfield96/PansimNuc: A nucleotide-level pangenome simulator.

A nucleotide-level pangenome simulator. Contribute to samhorsfield96/PansimNuc development by creating an account on GitHub.

github.com

📢 Deadline extended! The application deadline for our MOPITAS Autumn School on Spatial Transcriptomics Data Analysis, taking place October 12-14th in Copenhagen, has been pushed to August 16th. This is a great chance to sharpen your expertise in data science and multi-omics analysis. Register now!

Bild

We've been looking at how to compare and cluster large numbers of genomes, such as those in large isolate databases such as AllTheBacteria, and metagenome assemblies (e.g. SPIRE, MGnify). On a combined dataset of 5.6 million assemblies, we can now cluster/dereplicate everything in under a day!

Johanna von Wachsmann@johannavw.bsky.social · 3mo ago

🧬 New preprint! We clustered 5.6 million bacterial genomes into genomically cohesive units (GCUs) 500× faster than existing tools. (In just 14 hours, 16.5 GB RAM using 48 CPUs). 🦠🐙Meet gemsparcl 💎✨! www.biorxiv.org/content/10.6...

Disrupting phage liquid crystalline droplets restores antibiotic susceptibility in Pseudomonas aeruginosa biofilms out in @plosbiology.org by @abultarafder.bsky.social and team. Exciting collaboration with @geiselbiofilm.bsky.social @pearce-maths.bsky.social and others

PLOS Biology@plosbiology.org · 4mo ago

#Biofilm matrices containing filamentous phages help #Pseudomonas aeruginosa tolerate antibiotics. @abultarafder.bsky.social @tbharat-lab.bsky.social &co show that #nanobody disruption of #phage Pf4 #LiquidCrystalline droplets restores #antibiotic susceptibility @plosbiology.org 🧪 plos.io/4xkd6Mw

Top: Nanobody binders are potent inhibitors of Pf4 liquid crystalline droplet formation and disrupt preformed droplets. Cryo-ET of Pf4 liquid crystalline droplets incubated with Nb43. Tomographic slice of a Pf4 liquid crystalline droplet specimen incubated with (left) 0.1 μM and (right) 1 μM Nb43. Bottom: Schematic representation of nanobody action in abolishing antibiotic tolerance of P. aeruginosa biofilms. In untreated biofilms (left), cells show increased antibiotic tolerance due to Pf4 liquid crystalline droplets formed by depletion attraction in the biofilm EPS matrix, where encapsulated cells are protected by an antibiotic diffusion block. In nanobody treated biofilms, patchy binding of nanobody to Pf4 filaments reduces depletion attraction between the filaments preventing liquid crystalline droplet formation and encapsulation of cells, leading to increased antibiotic susceptibility of bacteria within the biofilm.

If you do any work with DNA sequencing at some point, you need to assemble the reads. Usually, we just use the contigs, but that ignores important evidence from your data, so @vijinim.bsky.social wrote agtools to explore the assembly graph and learn more! academic.oup.com/bioinformati...

agtools: a software framework to manipulate assembly graphs

AbstractMotivation. Assembly graphs are a fundamental data structure used by genome and metagenome assemblers to represent sequences and their overlap info

academic.oup.com

New work on using transformers (using gene order) for tasks in genomic epidemiology: www.biorxiv.org/content/10.6... We trained BART models (w/ extended context windows) on E.coli and S.pneumo from AllTheBacteria and tested if the model could find new strains, insertions (blaCTX-M) and co-selection

Embeddings of gene order from PanBART model
Sam Horsfield@samuelhorsfield.bsky.social · 5mo ago

Our new preprint is out! We train a transformer on gene order and gene content of bacterial pathogens, applying it to a range of epidemiological and evolutionary analyses (1/8) www.biorxiv.org/content/10.6...

I did my PhD on bacterial meningitis, finding whether there are genetic factors which make meningitis more likely. Wrote down some initial thoughts on the current outbreak in Kent: johnlees.me/posts/menb-o... I tried to think of factors and their likelihood to explain why this is happening now

menB outbreak in Kent -- initial thoughts

A cluster of meningitis cases is currently dominating the UK news. At least five of the cases have been confirmed as serotype B (menB), from what I can tell the others have not yet been analysed. My P...

johnlees.me

Working on trans-eQTL mapping with population-scale scRNA-seq? Meet LIVI, our latest framework that enables efficient trans-eQTL mapping at single-cell resolution across cohorts of hundreds to thousands of donors. Preprint: doi.org/10.64898/202... Summary in the 🧵 below by @danaivagiaki.bsky.social

doi.org

Danai Vagiaki@danaivagiaki.bsky.social · 8mo ago

Delighted to present Latent Interaction Variational Inference (LIVI), a framework for trans-eQTL mapping at single-cell resolution that I developed during my PhD together with colleagues from @steglelab.bsky.social 1/n

How do bacterial pangenomes evolve, what controls their dynamics, why do they exist? Fitting a mechanistic model to 450 species from allthebacteria.org suggesting fast vs slow gene exchange (i.e. amount of MGEs) is a major differentiating factor, correlated with phylogeny rather than lifestyle

Proportion of fast genes, as a trait mapped on the bacteria-wide phylogeny
Sam Horsfield@samuelhorsfield.bsky.social · 8mo ago

At long last, my final PhD chapter is out: we developed a novel evolutionary simulator of bacterial pangenomes, Pansim, fitting it to data from >600K genomes using a likelihood-free framework, PopPUNK-mod, to explore neutral and adaptive pangenome dynamics www.biorxiv.org/content/10.6...

Super excited to announce the release of gene and intergenic region annotation from the largest bacterial genome and MAG datasets available, including AllTheBacteria, GTDB, SPIRE, HRGM, mOTUs and MGnify - dereplicated and available from HuggingFace huggingface.co/AllTheBacteria

Hugging Face – The AI community building the future.

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co

Official Statement: Cambridge University, Department of Veterinary Medicine, 12/12/2025 "The decision to recommend the closure of what the Times Higher Education Supplement ranks as the best undergraduate veterinary course in the world has come as a bolt from the blue." [1] #SaveTheVetSchool

Really pleased to share the first paper to come out of the lab. We found that hospital patients were frequently colonised with P. aeruginosa and that the same clone was shared between the gut and the lung. The phylogenies indicate that the clones moved from lung->gut www.nature.com/articles/s41...

High frequency body site translocation of nosocomial Pseudomonas aeruginosa - Nature Communications

Here, the authors report within-host diversity and body site translocation dynamics in hospital samples of Pseudomonas aeruginosa and reveal that body site sharing was likely due to within-patient tra...

nature.com