The goal: move from statics to dynamics; from asking "what is a GRN made of?" to "how does a GRN map inputs to outputs?"
The challenge: Bridging data abundance & conceptual understanding
New experimental & computational approaches are needed
Read the details here:
www.nature.com/articles/s41...
We suggest 3 linked principles to build these models:
(1) models grounded in cellular & evolutionary biology
(2) molecular constraints to reduce the solution space
(3) synthetic biology & cis-regulatory perturbations to generate training data & test predictions
We argue for a return to causal GRN models
Using representation learning to capture the emergent logic of gene regulation without needing to model every molecular interaction
Think Marr's levels of analysis, applied to developmental biology
Single-cell genomics has given us unprecedented data to analyse GRNs
But more data ≠ more understanding
Current inference methods often perform no better than random guesswork & multimodal approaches bring new problems as fast as they solve old ones
New Perspective form Rory Maizels & me: "Gene regulatory networks: from correlative models to causal explanations"
Gene regulatory networks are supposed to give us mechanistic explanations of development, so why are we drowning in 'hairballs' of statistical correlations?
rdcu.be/e7zx7
It’s such shocking sad news we have lost Alan. He was such fun and inspiring on the CoB board where we met every 3 month for many years. So many great memories of Alan that I will always cherish.
06.03.2026 19:44 — 👍 10 🔁 2 💬 0 📌 0
New preprint from Meri Saez & co
How much information does a developmental landscape encode?
Integrating information theory, dynamical systems & dev bio to quantify the number of signalling regimes that generate distinguishable cell fates
biorxiv.org/content/10.64898/2026.03.03.709461v1
Interested in helping your community find the most interesting #preprints?
Apply to join our academic team at Development as a new Preprint Editor.
doi.org/10.1242/dev....
New from @dev-journal.bsky.social: we're launching a new initiative and hiring Preprint Editors to help navigate the growing world of preprints in developmental & stem cell biology
Join our community & shape the future of research
journals.biologists.com/dev/article/...
Apply to become a Preprint Editor Development
We are inviting experienced researchers to apply to become Preprint Editors, a new type of editorial role responsible for handling our 'In preprints' articles and forging links with preprint literature. #prepints
Learn more and apply before 30 March 2026:
journals.biologists.com/dev/pages/pr...
Are you a clinician looking for a fully-funded postdoctoral research opportunity?
Check out our postdoctoral career development fellowships for clinicians, and see how you can join our vibrant community: www.crick.ac.uk/careers-and-...
New from @dev-journal.bsky.social: we're launching a new initiative and hiring Preprint Editors to help navigate the growing world of preprints in developmental & stem cell biology
Join our community & shape the future of research
journals.biologists.com/dev/article/...
His Royal Highness Prince William is talking with Edith Heard, who is to his left, inside a well-lit building entrance. They are both smiling and walking towards the camera.
His Royal Highness Prince William Is sitting at a round table with a group of students. They are all wearing safety glasses, and he is holding a pipette. There is a variety of scientific instruments in the middle of the table.
His Royal Highness Prince William is sitting in front of a microscope wearing a lab coat. He is laughing while looking towards someone to his left, also in a lab coat. They are both inside a large white room.
His Royal Highness Prince William stands next two Edith heard, both with arms folded across their front. He is facing and talking to three other people inside a well-lit building entrance.
Thanks to His Royal Highness Prince William for visiting the Crick today to explore how education, research and innovation are helping to deliver benefits in the UK and across the world – from antimicrobial resistance to the biology underlying our mental health.
🔗 www.crick.ac.uk/news/2026-02...
James Briscoe Editor-in-Chief , Development The Francis Crick Institute, UK “Community journals such as Development play a crucial role. Every issue of Development has papers handled by academic editors who are leaders in the field, so it offers a curated collection of the latest developmental biology research selected by experts.” Key topics: cell fate control and differentiation computational biology and modelling gene regulation and stem cells neural development and patterning tissue engineering and organoids
Our Editor-in-Chief, @jamesbriscoe.bsky.social, will be speaking at 2nd Meeting on Early Embryogenesis and Epigenetics at @molgen.mpg.de from Monday 16 February.
Please speak to James about publishing your research in Development.
journals.biologists.com/dev/pages/aims
Very sad news. Gail Martin (1944-2026) was a figure in developmental biology. She pioneered the field of ES cells.
A great colleague, a friend and a lovely person.
For International Day of Women and Girls in Science, #IDWGS2026, we are featuring some of our 'Transitions in development' interviewees from the past year.
Browse our full series here: journals.biologists.com/dev/collecti...
Congratulations to Mina Gouti on her promotion to Professor
www.mdc-berlin.de/news/news/pr...
Our new paper: Entropy Sorting Feature Selection (ESFS)
A computational framework for gene selection from single cell data that extracts biological signals in noisy data while avoiding artefacts from conventional dimensionality reduction
A thread
www.biorxiv.org/content/10.6...
Fill out our community survey to help shape the future of JCS Journal of Cell Science logo Line drawing of a group of people with speech bubbles above their heads
After 15 years as Editor-in-chief (EiC), Michael Way @drmichaelway.bsky.social will be stepping down at the end of 2026 & we are seeking feedback from the cell biology community as part of the consultation process for appointing a new EiC.
We'd love to hear from you: www.surveymonkey.com/r/GZCC9F9
ESFS is available as open-source packages for both CPU and GPU
github.com/aradley/ESFS
All credit to Arthur Radley who led the project, supported by Giulia Boezio, Cameron Shand, & Ruben Perez-Carrasco
4. Neural tube development: ESFS disentangled three overlapping regulatory programmes (spatial, temporal, neurogenesis) and identified superior marker genes
03.02.2026 18:44 — 👍 1 🔁 1 💬 1 📌 0
3. Glioblastoma tumour spatial data (15 heterogeneous samples): ESFS identified global, shared & tumour specific expression programmes in a single workflow, no manual sample-by-sample annotation needed
03.02.2026 18:44 — 👍 0 🔁 0 💬 1 📌 0
2. Colon spatial transcriptomics: ESFS resolved Serosa–LP–IEC stratification throughout the entire proximal-distal axis and uncovered gene expression gradients that consensus NMF analysis had missed
03.02.2026 18:44 — 👍 1 🔁 0 💬 1 📌 0
We tested ESFS in 4 challenging datasets
1. Early human embryo development from 8 independent studies. Without any batch integration, ESFS revealed coherent developmental trajectories & identified previously uncharacterised trophectoderm subpopulations
The framework comprises three algorithms
ES-GSS separates technical noise from biological signals
ES-CCF identifies cell groupings that maximise correlation with individual genes
ES-FMG identifies minimal marker gene sets capturing distinct expression programmes without redundancy
ESFS uses information-theoretic measures to identify multivariate gene relationships directly in gene expression space, not in latent space
By reformulating conditional entropy as a sorting problem, it achieves tractable feature correlation quantification with built-in significance testing
The problem: HVG selection shows poor reproducibility across implementations and introduces systematic biases that affect downstream analyses
Yet while other analytical steps get sophisticated new tools, feature selection has been largely neglected
Single-cell RNA-seq has revolutionised biology, but little attention has been paid to the first step in the analysis: feature selection
We've seen huge advances in clustering & integration, but highly variable genes (HVG) is still the method used to start most analyses
Our new paper: Entropy Sorting Feature Selection (ESFS)
A computational framework for gene selection from single cell data that extracts biological signals in noisy data while avoiding artefacts from conventional dimensionality reduction
A thread
www.biorxiv.org/content/10.6...
I watched much of Ian Chapman's (Head of @ukri.org) appearance this morning at @ukparliament.parliament.uk Science, Innovation and Technology Committee..
A few points and new information that bioscientists may find interesting...
A 🧵
