My Mom would often take my brother and I into the lab with her... We practically grew up in the University. Not sure she thought about it as defiance, just necessity!

A Simple Act of Defiance Can Improve Science for Women nytimes.com/2024/04/26/opi…

Most interesting #cancer research work I have read in a long time. Cancer without mutation. Turn off PRC for 2 days, and JAK-STAT, Zeb1 get flipped on...

nature.com/articles/s4158…

Looking to support applications for this to come to our lab The ICR!

❤️❤️❤️❤️❤️❤️❤️❤️❤️

The research, in collaboration with Imperial College London, has identified two genes that could potentially be targeted to stop melanoma skin cancer from spreading.

Read the paper in Cell Reports 👇cell.com/cell-reports/f…

📰NEWS: Cancer cells shapeshift to spread around the body, but how do they know what shape to choose?

BakalLabICR - Cancer Morphodynamics developed a new technique to see cells in 3D, to study how cells know where they are and what shape they should change into.

Read more👇
icr.ac.uk/news-archive/s…

This paper was a long slog, so I'm thrilled to see it published. We used high-throughput light sheet microscopy to understand 3D #cellshape determination during #melanoma invasion. Credit to Lucas Dent for shepherding this work across the desert.
sciencedirect.com/science/articl…

🇨🇦 Who is #TerryFox ? And what is his incredible legacy?

We share his story and how he's inspired Canadian cancer researchers like the ICR's Professor Chris Bakal (BakalLabICR - Cancer Morphodynamics).

Find out more about Terry and sign up to the #TerryFox Run UK👇

icr.ac.uk/blogs/science-…

In Boston for #Ferment2024 . Let me know if anyone is here and keen to link up 🧬

🥇 Congratulations to Matt de Vries, the winner of our 'Early-career Entrepreneur of the Year' award, for his creation of Sentinal 4D and for inspiring an #entrepreneurial attitude.

#IEAwards24 #OnMyHorizon

biorxiv.org/content/10.110…

One explanation why i have problems getting grants supporting our #AI #cancer cell work these days...

'Our findings show that consensus about an idea’s worth diminishes the newer it is....'
nature.com/articles/s4156…

Right now most of #cancer #pathology is done in 2D. What if we could do it in #3D ? I think #diagnosis will get better and improve #patient outcomes. (I can't wait to let our #AI analyze these #3D images of #melanoma )

It's been the time for 25 years, but most people won't get this

'It’s time to admit that genes are not the blueprint for life' nature.com/articles/d4158…

Humans still surpass #AI at diagnosing #cancer using #images . That because many AI models focus on high-res analysis, missing the forest for the trees. However, we've developed an AI that mimics human examination by zooming in and out on cancer cells.
arxiv.org/abs/2305.05314

Was I happy with the answer?

Our work implementing Multiple Instance Learning #MIL #AI models for more accurate #cancer #pathology . Very excited because these models consider aspects of spatial organization and phenotypic #heterogeneity (which most others don't do!)
arxiv.org/abs/2305.05314

Nature research paper: Discovery of a structural class of antibiotics with explainable deep learning go.nature.com/41w3TkE

The classic dinosaur.

To big directions we are looking for 2024 is #explainableAI for #digitalpathology , and #voltage imaging of #cancer . My collaborators Mustafa Djamgoz, Amanda Foust, Peter Quicke, and Yilin Sun put together this review on #voltage imaging. Imperial College London doi.org/10.1089/bioe.2…