Lightnews — Scholar-powered news

Reposted by Hugh Selway-Clarke

Calum Gabbutt @calumgabbutt.bsky.social · 28d

Cancer is an evolutionary disease, but does knowing a cancer’s evolutionary past help predict its future? Out today in @nature, we learnt the evolution of 2000 lymphoid cancers and found it was highly correlated with clinical outcomes! (1/7)
rdcu.be/eFrrc

Fluctuating DNA methylation tracks cancer evolution at clinical scale

Nature - Cancer evolutionary dynamics are quantitatively inferred using a method, EVOFLUx, applied to fluctuating DNA methylation.

rdcu.be

1 19 46

Reposted by Hugh Selway-Clarke

Trevor Graham @trevorgraham.bsky.social · 28d

Studying cancer evolution needs multi-region or single cell seq for phylogenetics, right? Amazingly (I think!) we found single-sample bulk methylation suffices, via analysis of "fluctuating methylation". In @nature.com today led by brilliant @calumgabbutt.bsky.social www.nature.com/articles/s41...

Fluctuating DNA methylation tracks cancer evolution at clinical scale - Nature

Cancer evolutionary dynamics are quantitatively inferred using a method, EVOFLUx, applied to fluctuating DNA methylation.

www.nature.com

7 39 92

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

I’m grateful to my co-authors and the Janes lab @lungs4living.bsky.social for their support over the last 4 years, as well as the patient donors and clinical teams on whose data this work relies.
If you have any questions about the pre-print, do reach out! doi.org/10.1101/2025...
[15/15]

Recovery of human upper airway epithelium after smoking cessation is driven by a slow-cycling stem cell population and immune surveillance

The upper airway epithelium in humans is maintained in homeostasis by a resident population of basal stem cells. In the presence of tobacco smoke these gain mutations that significantly increase their...

doi.org

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

Finally, this model is a rapid perturbable system for the accumulation of genomic damage in the presence of carcinogens and immune modulation. This can be applied for in silico trials of preventative therapeutics, or to other cancers with a similarly dominant causal mutagen.
[14/15]

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

Lots more can and should be done to further validate this picture: spatially resolved data would give us another viewpoint, and direct in vivo and in vitro interventional studies are, as ever, crucial. Co-evolution of lung and immune cells would be likely here and could be measured!
[13/15]

1 1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

So what does this mean for our understanding of the upper airway? These results paint a tentative picture, in which smoking both causes mutational damage and suppresses an immune mechanism of recovery. Quiescent cells, always present, then provide a competitor to take over after smoking.
[12/15]

Schematic showing the proposed paradigm of somatic evolution in the upper airway in response to smoke

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

Adapting these same classifiers to the observed scWGS data, we found that distinct machine learning classification methodologies converged on the same answer: a quiescent subpopulation of cells, and smoking-suppressed immune predation of mutated stem cells.
[11/15]

Barplot showing the frequency with which classifiers classify the observed dataset as each hypothesis combination. Logistic regression and random forest methodologies agree in classifying the true data as "Quiescent, Immune Response"

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

When tested on unseen simulations, the classifiers were very accurate on some problems (identifying a quiescent subpopulation of cells) and were pretty hopeless at others (changes to the fitness landscape in the presence of smoke), giving us a good idea of what we can learn from this data.
[10/15]

Heatmap showing the identifiability of the combination of hypotheses generating a simulation from the simulation outputs.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

An aside: these big sets of lifelong simulations rack up a long time in simulated somatic evolution - almost 70 million years. To run this in vivo and in series, we'd have to have started in the Cretaceous! A good advert for efficient code and in silico modelling...
[9/15]

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

Qualitative fits are nice, but with this data and this modelling approach we can go further. We created a collection of synthetic datasets from literature-informed priors, and trained machine learning classifiers to identify which hypotheses were active in a particular dataset.
[8/15]

Schematic showing the inference structure: simulated cohorts are generated with prior-sampled parameter values in each combination of hypotheses. These are used to train machine learning classification methods to learn the combination of hypotheses. These can then be applied to the observed dataset.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

First, we had to check that these hypotheses could reproduce the surprising findings: they could! Different combinations of hypotheses looked qualitatively similar to what we saw in the scWGS data, in quite different ways (spatial clonal patterning completely different between paradigms!)
[7/15]

Distributions of mutational burden (shown as distributions and as spatial heat maps) within two distinct combinations of hypotheses: Protected subpopulation, immune response and Quiescent subpopulation, smoking-addicted drivers

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

We modelled this on the natural 2D spatial structure of the upper airway epithelium. We added in each mechanistic hypothesis modularly, so that we could run simulations with any combination of them.
[6/15]

Spatial lattice showing the implementation of the model.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

To answer these questions, we implemented an established model for this stem cell population to simulate a small population of lung stem cells for each patient over their entire lifetimes, based on their smoking histories.
[5/15]

Diagram showing an established model of somatic evolution via fitness-engendered changes to the cell fate distribution.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

We collated a set of mechanistic hypotheses about the upper airway epithelium that could explain these dynamics. It seems reasonable that any of these could lead to the observed dynamics, but is it true? And which of them is most likely given the observed data?
[4/15]

Schematic showing the structure of the upper airway epithelium, and mechanistic hypotheses to explain the surprising observations in the scWGS data.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

Where it gets interesting is in the variation within an individual's lung stem cells. Some cells from lifelong smokers look like they come from a never-smoker’s lungs, and there’s more of these less-mutated cells in the lungs of ex-smokers than ongoing smokers!
[3/15]

Mutational burden distributions within the lungs of one never-smoker, one smoker and one former smoker. The never smoker has a tight distribution around 2000 mutations. The smoker has a broader distribution closer to 10,000 mutations, with a small extra population around 2000. The former smoker has a population around 7000 and an extra population around 2000, forming a larger proportion of the cells than in the smoker.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

This work started with two surprising observations in recently published single cell-derived whole genome sequencing (scWGS) data from stem cells in the upper airways of patient donors. As you'd expect, the data showed an accumulation of mutations over lifetime, accelerated by smoking.
[2/15]

Graph showing mutational burden against age. Mutations accumulate with age, faster for ever-smokers.

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Sep 3

My main PhD project is now up on bioRxiv! If you’re interested in somatic evolution or machine learning for simulation-based inference, read on… 
doi.org/10.1101/2025...
[1/15]

Recovery of human upper airway epithelium after smoking cessation is driven by a slow-cycling stem cell population and immune surveillance

The upper airway epithelium in humans is maintained in homeostasis by a resident population of basal stem cells. In the presence of tobacco smoke these gain mutations that significantly increase their...

doi.org

1 3 8

Reposted by Hugh Selway-Clarke

Charlotte Capitanchik @tinycaptain.bsky.social · Sep 2

Had such a great time at #Londonomics2025 today - excellent talks and discussions - thank you to everyone who came along and shared their ideas with us.

1 2

Reposted by Hugh Selway-Clarke

James Reading @jamesreading.bsky.social · Aug 5

🚨 PhD Position available in our lab 🚨 exploring the power of blood immune multi-omics to detect lung cancer years prior to clinical diagnosis in a unique cohort of >10,000 CT screened individuals.
✅ Wet & dry lab
✅ September 2025 enrolment
✅ UK tuition fees only

www.ucl.ac.uk/medical-scie...

Pre-Cancer Immunology

The Pre-Cancer Immunology Lab (James Reading Lab) is mapping pre-invasive T cell dynamics during carcinogenesis to detect and intercept cancer development.

www.ucl.ac.uk

15 16

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Jul 29

📣📣 Another Londonomics Symposium! 📣📣

Last year was a great mix of computational biologists from biology, maths and computing backgrounds - I enjoyed it so much I joined the organising committee! Excited to meet people and talk cool science (including AlphaGenome, keynote from @avsecz.bsky.social!)

londonomics.bsky.social @londonomics.bsky.social · Jul 29

Are you an Early Career Researcher in bioinformatics? Then this symposium is for you 💡

Join us for a day of talks, networking and career discussions. Present your work to get fresh new ideas and the chance to win prizes 💸

Featuring @avsecz.bsky.social of Google DeepMind as our keynote speaker⚡️

1

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Jul 26

A day @ewanbirney.bsky.social posts a thread of "some musings" is a good day - I haven't (yet!) used MR in anger but this is a great checklist for how to do it properly.

Ewan Birney @ewanbirney.bsky.social · Jul 26

Had to finish this thread in a rush without a concluding flourish but adding it here for my “how to do Mendelian Randomisation without messing up”

Ewan Birney @ewanbirney.bsky.social · Jul 26

Some musing on Mendelian Randomisation as a technique, triggered by discussions with a variety of colleagues. TL;DR Mendelian Randomisation works well when it used for hypothesis testing of valid exposure => outcome scenarios, as long as the MR assumptions hold (obvs!) and it is performed with care.

Reposted by Hugh Selway-Clarke

Sophie de Carné (she/her) @sdecarne.bsky.social · Jul 22

Writing is thinking. A reminder on why scientists write, and why we should continue writing in the era of large-language models by Nature Review Bioengineering

@natureportfolio.nature.com
www.nature.com/articles/s44...
#AI #LLM 🧪

Writing is thinking - Nature Reviews Bioengineering

On the value of human-generated scientific writing in the age of large-language models.

www.nature.com

1 7 33

Hugh Selway-Clarke @hughselwayclarke.bsky.social · May 2

Great to see this work from @sandra-gl.bsky.social published yesterday, using mouse models to dig into the early stages of lung squamous cell carcinoma. Beautiful work on an important disease, congratulations to all involved!

lungs4living.bsky.social @lungs4living.bsky.social · May 2

Sandra's paper is out in Science! So pleased. Thank you to all the authors, contributors and patients that have been with us on this journey. @science.org @sandra-gl.bsky.social Aberrant basal cell clonal dynamics shape early lung carcinogenesis
www.science.org/doi/full/10....

www.science.org

1

Reposted by Hugh Selway-Clarke

lungs4living.bsky.social @lungs4living.bsky.social · Mar 26

Congratulation to the SUMMIT team on our publication in Lancet Oncology today: Low-dose CT for lung cancer screening in a high-risk population (SUMMIT): a prospective, longitudinal cohort study.
authors.elsevier.com/sd/article/S...

1 2 4

Hugh Selway-Clarke @hughselwayclarke.bsky.social · Mar 21

Congratulations Alex!! 🥳

1