
Jewellery designer and number one fan of @DrBPChamberlain
London based. Yorkshire born.
Joined October 2013
- Tweets923
- Following229
- Followers131
- Likes517
Annabel Chamberlain retweeted
Off to @NeurIPSConf next week? So are we! 😎 Come meet our team @WillMcCorki1
@_joanna_yoo @john_pryan @_lychrel Explore exciting opportunities in #AI-powered drug discovery for ML researchers and engineers. Join our mission to conquer disease. 🔬💪tinyurl.com/ycx7dt3a
Annabel Chamberlain retweeted
Interested in what we're doing at @IsomorphicLabs?
This is the most we've shared so far. Needs a login - but it's free...
... and we're hiring
Thank you to @AndrewE_Dunn from @endpts for the in-depth piece on @IsomorphicLabs, exploring the tech we’re developing to reimagine drug discovery using AI, and the passion of the team leading our work to achieve Iso’s mission.
In the feature, Andy tracks the timeline of ‘The meteoric rise of Isomorphic Labs (and AI)’ and provides ‘a glimpse behind the curtain into how Hassabis’ AI-driven vision is becoming reality, and why the company believes its pure focus on technology and ideas will lead to success compared with companies taking different approaches.’
Read more here: endpts.com/isomorphic-labs-c…
Annabel Chamberlain retweeted
Entering my 3rd year in drug discovery after leaving ML research at Twitter. I'm sharing the BEST resources (books, courses, tools) that helped me make the jump 👇
First up: BOOKS that changed the game for me:
General Bio/Science:
The Double Helix, James Watson - A classic.
What is Life? by Nobel laureate and @IsomorphicLabs advisor Paul Nurse.
The Code Breaker, bio of CRISPR starring Nobel laureate and @IsomorphicLabs advisor Jennifer Doudna by @WalterIsaacson,
Lessons in Chemistry, Bonnie Garmus. My wife got me this as a joke. Not what I expected. Great book though.
Cancer Focused:
The Emperor of All Maladies, Siddartha Mukherjee
The Breakthrough @TheGoodNurseBK
The Death of Cancer @DeVitaDoctor
The Cancer Code @drjasonfung
Drug Discovery Specific:
For Blood and Money, the development of BTK inhibitors, @nathanvardi
The Billion-Dollar Molecule, the Vertex story, Barry Werth
Textbooks (for the hardcore):
Medicinal Chemistry 7th Edition. Graham L Patrick. For how to drug different protein classes.
Molecular Biology of the Cell, 5th edition, Bruce Alberts.
What are YOUR go-to resources? Let's swap recommendations! #DrugDiscovery #ScienceTwitter #BioTech #ML Courses and other tricks and tools to follow…
Annabel Chamberlain retweeted
Intrigued by drug discovery? 🧪 This article on the rise & fall of @reverie by former CEO @agupta is a MUST-READ! 🤯
Key takeaways:
1. Big Pharma don’t like to buy software.
2. Bio $ for molecules? 💸 Tread carefully...
3. Develop your own chemical matter
4. Even with top tech, odds are low. 🍀
Agree? Disagree? Let's discuss! 👇 #drugdiscovery #biotech #startups #machinelearning
With the start of my first batch as a visiting partner at @ycombinator, this felt like a good time to write my founder story about our journey at @reverielabs. It's a bit of a long one.
ankitg.me/blog/2025/01/06/un…
Annabel Chamberlain retweeted
Really cool intro to computational protein folding featuring some current and former colleagues...
...and if you interested in the physics of proteins folding, then lecture 5 from @eriklindahl youtube.com/playlist?list=PL… is a great place to go.
#drugdiscovery #AI4Science
Annabel Chamberlain retweeted
2 years ago, I left ML research at Twitter to dive into drug discovery—a field I knew nothing about. 💡
I last studied biology/chemistry at 16. I previously shared by top books. Now here are the FREE online courses that taught me the fundamentals - rated by how much they helped me pivot careers. 🧬💻
Biology
@edXOnline Introduction to Biology - The Secret of Life. If you do just one, do this one. @eric_lander is a great teacher, who features in The Code Breaker, the story of CRISPR that made my top books 5/5
Chemistry (@MITxCourses @ChemistryMIT @edx)
General Chemistry I. Chemical bonding and molecular geometry. Essential. 5/5
General Chemistry II: Gibbs free energies, acid and bases, reaction rates and enzymes. Matt Shoulders - top lecturer, great name. 4/5
Biochemistry (do after biology & chemistry. Obvs)
HarvardX @RachelleGaudet: Principles of Biochemistry. More advanced. Amino acids, protein structure, Kreb cycle and metabolism. 3/5.
@MITxCourses @edXOnline Biochemistry: Biomolecules, Methods, and Mechanisms
Chemical biology. Next on my list.
Drug discovery
@H3D_UCT @coursera: Introduction to Small Molecule Drug Discovery & Development. Terminology and pipeline of drug discovery. Inspiring w/ a perspective on infectious diseases and traditional medicine in Africa. 4/5
Cancer
@coursera @JohnsHopkins three courses on cancer. Made in 2016 and lower quality than the others here. 2/5
What are YOUR go-to resources? What did I miss? Let's swap recommendations! #DrugDiscovery #ScienceTwitter #BioTech #ML Next up the essential tricks and tools…
Annabel Chamberlain retweeted
🚀 From ML at Twitter to Drug Discovery: Practical Tips That Made a REAL Difference! 🔬💡
Two years ago, I made the leap from ML research to drug discovery. I’ve shared books & courses before—now it’s time for hands-on tools & tricks that helped me level up. Let’s dive in! 👇
GenAI for Science
Like everyone, I use LLMs daily. But these two tools have raised the bar:
🔥 Gemini Deep Research – Need a crash course on a new topic?
🔍 Example: “Summarize the latest cancer therapies, including effectiveness data.”
📄 Output? A 9-page report, 34 references, & insights from 94 sources—auto-magically.
📚 NotebookLM – Your personal AI-powered research assistant.
I uploaded:
📖 Medicinal Chemistry – Graham L. Patrick
📖 Drug Design – Gerhard Klebe
📖 Molecular Biology of the Cell – Alberts
Now, I have a chatbot trained on my books—with citations! 🔥
Bonus? It turns notes into custom podcasts 🎙️. Want a deep dive on drugging ion channels? Just ask. Perfect for the morning 🚴.
Molymod: The OG 3D Visualizer
Can you mentally rotate a 3D protein with dozens of degrees of freedom? Neither can I. Nothing beats a physical Molymod model for grasping conformation & chirality. I keep a tripeptide on my desk.
What to Memorize (Yes, Even in the GenAI Era)
Turns out, chemists can talk faster than LLMs can think. What I committed to memory:
🧬 Amino acid structures, properties & one-letter codes
🧪 Common chemical functional groups & their suffixes (e.g., -ane, -ene, -zole)
💊 Key drug properties (logP, logD, Papp A-B, etc.)
The Power of Notes
Maybe future AI agents will store everything, but for now? I keep a Google Docs network of interlinked notes—a modern version of the commonplace book used by Da Vinci, Newton & Darwin.
💡 What are your BEST tips and tricks? What did I miss? Let’s swap recommendations! 👇 #DrugDiscovery #ScienceTwitter #BioTech #ML
Next up: Going the other way? Natural science to ML … stay tuned! 🚀
Hey @glossier. The label is coming off my skin tint and results in me getting white bits on my face like bad glitter. It's never happened before but it's quite annoying. I think you've got a dodgy batch.
306
Hey @Bodenclothing! I get a lot of leaflets from you to my house fir the previous owner. Please could you take me off your mailing list?
1
194
Well done, @DrBPChamberlain and team! 🥳🎉🍾 I'm happy all your hard work paid off. See you in Kigali! #ICLR2023
1/3 Twitter graph ML full house at #ICLR2023 :
📜Poster arxiv.org/abs/2210.00513
🔦Spotlight arxiv.org/abs/2210.01542 top 25%
👄Oral arxiv.org/abs/2209.15486 top 5%
spanning my recent interests in continuous GNNs, hyperbolic geometry (applied to RL this time) & link prediction
7
824
1/4 Today we released two new link prediction models developed over the past 12m. We solve many problems that prevent GNNs succeeding at link prediction by adding a message passing mechanism that uses sketches of subgraphs.
arxiv.org/abs/2209.15486
1
Loved Project Dictator ⭐⭐⭐⭐⭐
Hilarious, thought provoking and weirdly wonderful. The guys are fantastic.
#EdFringe2022 #EdFringe22 #edfringe #pleasancecourtyard
We have 2 shows up @edfringe this year!!! PROJECT DICTATOR at 6.10pm & our show for kids EVERYTHING HAS CHANGED at 10.40am both @ThePleasance courtyard. tickets.edfringe.com/whats-o…
tickets.edfringe.com/whats-o…
Loved this!⭐⭐⭐⭐⭐Full of energy, mystery, laughs & superb choreography. Gives slight Scooby Doo vibes and had me gripped from start to finish. A very slick production well worth seeing. #edinburghfestival2022 #PleasanceDome #speeddial #EdFringe22 #EdFringe2022 @ThePleasance
Our @edfringe trailer is here! 🎬
Don’t miss our brand new comedy-thriller #SPEEDDIAL coming to @ThePleasance this summer ☎️👀
🎟👉 pleasance.co.uk/event/speed-…
I am all for this.
Once again @gordic_aleksa doing a better job of describing our work than I can. From now on, instead of doing talks, I'm just going to beatbox over his videos
1
Hey @MinorFigures! Please ecluld someone come back to my email? I can't log in to my account and have had two deliveries in two days!
Annabel Chamberlain retweeted
3/3 if you REALLY can't wait, the gist of it is...
1/ We make over-squashing in GNNs & the related bottleneck phenomena precise
2/ Develop a new notion of graph curvature. Negatively curved edges => over-squashing
3/ Show that discrete Ricci flow can solve over-squashing
Annabel Chamberlain retweeted
2/3 Very grateful to have been part of this collaboration between Twitter & Oxford (w/ a dash of Imperial). Interested and can’t wait until ICLR to find out more? We will be presenting w/ Q&A in @HannesStaerk LOGAG reading group hannes-stark.com/LoGaG-readi… next Tuesday
Annabel Chamberlain retweeted
1/3 New paper showing that Ricci flow, best known as part of Perelman’s Poincaré Conjecture proof, can be used to rewire graphs & improve GNN performance. openreview.net/forum?id=7Umj… will be an oral (top 1.5%) at #ICLR2022
Thrilled that our paper "Understanding over-squashing and bottlenecks on graphs via curvature" has been accepted for an oral presentation at #ICLR2022! Big thank you to the team @Francesco_dgv @b_p_chamberlain @epomqo @mmbronstein (arxiv.org/abs/2111.14522)













