Sign in

John Lees

@johnlees.bacpop.org
1.6K followers 401 following 178 posts

Pathogen informatics and modelling bacpop.org (Research group leader at EMBL-EBI - ebi.ac.uk/research/lees)

PostsRepliesMedia
John Lees @johnlees.bacpop.org · 06/05/2026
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
13011
John Lees @johnlees.bacpop.org · 09/02/2026
This started as an opportunistic reuse of poppunk/sketchlib core-accessory distances, which we noticed differed between species in their shape (e.g. accessory distance at same core; faster/slower accumulation of accessory changes with core)
Core trees, core-accessory distance plots, and model fits for four bacterial species (A) Mycobacterium tuberculosis, (B) Streptococcus pneumoniae, (C) Escherichia coli and (D) Listeria monocytogenes.)
122
John Lees @johnlees.bacpop.org · 09/02/2026
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
17132
John Lees @johnlees.bacpop.org · 06/01/2026
Thank you for asking, yes Stubentiger does indeed stand for STUdying Balancing Evolution (Nfds) To Investigate GEnome Replacement
STUdying Balancing Evolution (Nfds) To Investigate GEnome Replacement
(also a German word for cat)
030
John Lees @johnlees.bacpop.org · 23/06/2025
I was lucky to go dog sledding earlier this year which, naturally, made me think about the Central Limit Theorem Quick posts about the CLT: www.johnlees.me/posts/dogs-c... (and some pictures of dogs)
Non-normal dogs, normal sledsPicking packs for sledsDraws of dogs
2113
John Lees @johnlees.bacpop.org · 30/04/2025
Pipeline overview for PopPIPE, with three different modes possible
021
John Lees @johnlees.bacpop.org · 04/04/2025
A familiar image of vaccine conjugation
020
John Lees @johnlees.bacpop.org · 04/04/2025
Trying to find an image of vaccine conjugation and google mostly coming up with the one from this paper www.mdpi.com/2076-393X/9/... Terrible resolution. I thought the font seemed really familiar too
Vaccine conjugation as inspired by an identical image from an earlier paper
130
John Lees @johnlees.bacpop.org · 10/01/2025
Why does this 'publisher' look like it is trying to sell me a zombie knife
.....Disseminate your Research or face the consequences
121
John Lees @johnlees.bacpop.org · 05/12/2024
Had a great time at the SMBE satellite meeting on pangenomes and pathogen transmission which we hosted in Tokyo last week! We had 87 on site attendees and 48 remote participants, with sessions on metagenomics, selection, emerging diseases, pangenomes and modelling.
SMBE 'japangenome' conference photoVegan goyzaA participant enjoying the whale at the venueA 300-yr old black pine tree
1140
John Lees @johnlees.bacpop.org · 25/09/2024
Now published in bioinformatics, CELEBRIMBOR 'pangenomes from metagenomes' academic.oup.com/bioinformati... I think reviewer comments helped make it better, so worth another look (at fig 1 particularly!)
5148
John Lees @johnlees.bacpop.org · 10/04/2024
Sam has then wrapped it all up in a nice snakemake pipeline CELEBRIMBOR (Core ELEment Bias Removal In Metagenome Binned ORthologs) which lets you go from MAGs to pangenome github.com/bacpop/CELEB...
110
John Lees @johnlees.bacpop.org · 10/04/2024
Here, Joel instead made a probabilistic model to use these to readjust gene frequencies to make them more accurate, allowing parameterised choices of how to do so, in the package core gene threshold (cgt) github.com/bacpop/cgt
111
John Lees @johnlees.bacpop.org · 13/02/2024
It's called GNASTy (Graph-based Nanopore Adaptive Sampling Typing, pronounced ‘nasty’)
It's Gnasty Gnorc
120
John Lees @johnlees.bacpop.org · 23/12/2023
Here is my yearly writeup/review of video games from the past year www.johnlees.me/posts/review... Best one: Chants of Sennar
210
John Lees @johnlees.bacpop.org · 12/12/2023
New blog post I gave a talk ~4yrs ago 'Multiple horses for multiple courses' about interfacing multiple programming languages for research software (particuarly python and C++). I recap that here, and end with some reflections on following my own advice: www.bacpop.org/blog/horses/
Multiple horses for multiple courses
000
John Lees @johnlees.bacpop.org · 17/10/2023
Hello new platform, please see my recent adverts Postdoc ad currently open: www.embl.org/jobs/positio... Recent paper: genome.cshlp.org/content/earl... www.ebi.ac.uk/about/news/r...
1135