Chaitanya K. Joshi
@chaitjo.bsky.social
AI researcher excited about biomolecule design π§¬
PhD student at the University of Cambridge
Prev. at FAIR, Prescient Design, and MRC LMB
π https://chaitjo.substack.com
PhD student at the University of Cambridge
Prev. at FAIR, Prescient Design, and MRC LMB
π https://chaitjo.substack.com
Pinned
Chaitanya K. Joshi
@chaitjo.bsky.social
Β· Dec 3
Introducing gRNAde: our own little "AlphaGo Moment" for RNA design! π§¬π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
I recently joined The Pauling Principle podcast to discuss our research on RNA design, hot takes on architecture research in BioML, origins of life, and being in a wet lab!
Feels pretty surreal that I'm interesting enough to be on a podcast (?)
www.youtube.com/watch?v=ftrQ...
Feels pretty surreal that I'm interesting enough to be on a podcast (?)
www.youtube.com/watch?v=ftrQ...
February 6, 2026 at 8:47 AM
I recently joined The Pauling Principle podcast to discuss our research on RNA design, hot takes on architecture research in BioML, origins of life, and being in a wet lab!
Feels pretty surreal that I'm interesting enough to be on a podcast (?)
www.youtube.com/watch?v=ftrQ...
Feels pretty surreal that I'm interesting enough to be on a podcast (?)
www.youtube.com/watch?v=ftrQ...
Four years of research on Geometric Deep Learning & Molecular Design, summarized in 3 questions:
- How expressive are our models? (Theory)
- Can we build foundation models for molecules? (Architecture)
- Can AI design RNA in the real world? (Application)
- How expressive are our models? (Theory)
- Can we build foundation models for molecules? (Architecture)
- Can AI design RNA in the real world? (Application)
January 8, 2026 at 4:38 AM
Four years of research on Geometric Deep Learning & Molecular Design, summarized in 3 questions:
- How expressive are our models? (Theory)
- Can we build foundation models for molecules? (Architecture)
- Can AI design RNA in the real world? (Application)
- How expressive are our models? (Theory)
- Can we build foundation models for molecules? (Architecture)
- Can AI design RNA in the real world? (Application)
New blog π: I reflect on why I worked on what I worked on...
I think a PhD is a very special time. You get to challenge yourself, push your boundaries, and grow. My thoughts go against the current AI/academia narrative online, so I hope you find it interesting.
chaitjo.substack.com/p/phd-thesis...
I think a PhD is a very special time. You get to challenge yourself, push your boundaries, and grow. My thoughts go against the current AI/academia narrative online, so I hope you find it interesting.
chaitjo.substack.com/p/phd-thesis...
A Cambridge PhD thesis in three research questions
Geometric Deep Learning for Molecular Modelling and Design: A personal scientific journey
chaitjo.substack.com
January 8, 2026 at 4:38 AM
New blog π: I reflect on why I worked on what I worked on...
I think a PhD is a very special time. You get to challenge yourself, push your boundaries, and grow. My thoughts go against the current AI/academia narrative online, so I hope you find it interesting.
chaitjo.substack.com/p/phd-thesis...
I think a PhD is a very special time. You get to challenge yourself, push your boundaries, and grow. My thoughts go against the current AI/academia narrative online, so I hope you find it interesting.
chaitjo.substack.com/p/phd-thesis...
It relies on new algorithms for template search, invented by Kagglers such as G. John Rao. If no templates are found, the model can still work and works decently well without templates, too.
January 4, 2026 at 1:43 AM
It relies on new algorithms for template search, invented by Kagglers such as G. John Rao. If no templates are found, the model can still work and works decently well without templates, too.
The main takeaway from the Kaggle competition was how powerful RNA template modeling can be β it was kind of shocking to us π€―
RNAPro integrates 3D templates into Protenix/the AF3 architecture, and this moves the needle on automated RNA structure prediction!
RNAPro integrates 3D templates into Protenix/the AF3 architecture, and this moves the needle on automated RNA structure prediction!
January 2, 2026 at 1:19 AM
The main takeaway from the Kaggle competition was how powerful RNA template modeling can be β it was kind of shocking to us π€―
RNAPro integrates 3D templates into Protenix/the AF3 architecture, and this moves the needle on automated RNA structure prediction!
RNAPro integrates 3D templates into Protenix/the AF3 architecture, and this moves the needle on automated RNA structure prediction!
Happy new year! A step change in RNA structure prediction, powered by top Kaggle-ers in a collaboration lead by Stanford University and NVIDIA
Happy to have played a small part in the new RNAPro model, significantly outperforming AlphaFold 3 as well as VFold (CASP winners)
Happy to have played a small part in the new RNAPro model, significantly outperforming AlphaFold 3 as well as VFold (CASP winners)
January 2, 2026 at 1:19 AM
Happy new year! A step change in RNA structure prediction, powered by top Kaggle-ers in a collaboration lead by Stanford University and NVIDIA
Happy to have played a small part in the new RNAPro model, significantly outperforming AlphaFold 3 as well as VFold (CASP winners)
Happy to have played a small part in the new RNAPro model, significantly outperforming AlphaFold 3 as well as VFold (CASP winners)
The video was largely created using AI text-to-video, and I must say the results are pretty impressive! There's huge potential for this technology to make science more approachable, just like Google's NotebookLM.
December 26, 2025 at 9:04 AM
The video was largely created using AI text-to-video, and I must say the results are pretty impressive! There's huge potential for this technology to make science more approachable, just like Google's NotebookLM.
Merry Christmas!ππ
Sharing something interesting: Friso van de Stadt created a pretty neat explainer video for our gRNAde paper and "AlphaGo moment for RNA design" blogpost!
www.youtube.com/watch?v=wDeZ...
Sharing something interesting: Friso van de Stadt created a pretty neat explainer video for our gRNAde paper and "AlphaGo moment for RNA design" blogpost!
www.youtube.com/watch?v=wDeZ...
AI vs. Humans: The "AlphaGo Moment" for RNA Design
YouTube video by Chaitanya K. Joshi
www.youtube.com
December 26, 2025 at 9:04 AM
Merry Christmas!ππ
Sharing something interesting: Friso van de Stadt created a pretty neat explainer video for our gRNAde paper and "AlphaGo moment for RNA design" blogpost!
www.youtube.com/watch?v=wDeZ...
Sharing something interesting: Friso van de Stadt created a pretty neat explainer video for our gRNAde paper and "AlphaGo moment for RNA design" blogpost!
www.youtube.com/watch?v=wDeZ...
I wrote some personal reflections about being physically embedded in a world-leading molecular biology lab @mrclmb.bsky.social, learning to communicate with experimentalists, back-breaking wet lab work, scientific rigour, and skin in the game!
π: chaitjo.substack.com/p/an-ai-rese...
π: chaitjo.substack.com/p/an-ai-rese...
An AI Researcher in the Cathedral of Molecular Biology
How we designed catalytic RNA functions, and what I learned holding a pipette at the MRC Laboratory of Molecular Biology.
chaitjo.substack.com
December 18, 2025 at 3:38 PM
I wrote some personal reflections about being physically embedded in a world-leading molecular biology lab @mrclmb.bsky.social, learning to communicate with experimentalists, back-breaking wet lab work, scientific rigour, and skin in the game!
π: chaitjo.substack.com/p/an-ai-rese...
π: chaitjo.substack.com/p/an-ai-rese...
- How to set up the design prompt just right?
- How many sequences to sample and how to make them diverse?
- How exactly is filtering done step-by-step?
- How to analyze wet-lab results and create figures?
You can find out all of that in our GitHub codebase!
π»: github.com/chaitjo/geom...
- How many sequences to sample and how to make them diverse?
- How exactly is filtering done step-by-step?
- How to analyze wet-lab results and create figures?
You can find out all of that in our GitHub codebase!
π»: github.com/chaitjo/geom...
GitHub - chaitjo/geometric-rna-design: gRNAde is a Generative AI framework for inverse design of 3D RNA structure and function
gRNAde is a Generative AI framework for inverse design of 3D RNA structure and function - chaitjo/geometric-rna-design
github.com
December 15, 2025 at 9:26 AM
- How to set up the design prompt just right?
- How many sequences to sample and how to make them diverse?
- How exactly is filtering done step-by-step?
- How to analyze wet-lab results and create figures?
You can find out all of that in our GitHub codebase!
π»: github.com/chaitjo/geom...
- How many sequences to sample and how to make them diverse?
- How exactly is filtering done step-by-step?
- How to analyze wet-lab results and create figures?
You can find out all of that in our GitHub codebase!
π»: github.com/chaitjo/geom...
With this release, we are fully openly releasing all the tacit knowledge for getting biomolecule design models to actually work in wet labs!
Seemingly small-but-critical details are often missing in flagship papers, and it took us a while to finetune our pipeline:
Seemingly small-but-critical details are often missing in flagship papers, and it took us a while to finetune our pipeline:
December 15, 2025 at 9:26 AM
With this release, we are fully openly releasing all the tacit knowledge for getting biomolecule design models to actually work in wet labs!
Seemingly small-but-critical details are often missing in flagship papers, and it took us a while to finetune our pipeline:
Seemingly small-but-critical details are often missing in flagship papers, and it took us a while to finetune our pipeline:
Many papers have built their RNA design pipelines using the gRNAde code & data from our ICLR 2025 spotlight paper - that's been exciting to see! Especially the attempt at biologically meaningful benchmarks π€
But how to take ML conference papers to wet-lab validated results?
But how to take ML conference papers to wet-lab validated results?
December 15, 2025 at 9:26 AM
Many papers have built their RNA design pipelines using the gRNAde code & data from our ICLR 2025 spotlight paper - that's been exciting to see! Especially the attempt at biologically meaningful benchmarks π€
But how to take ML conference papers to wet-lab validated results?
But how to take ML conference papers to wet-lab validated results?
Excited to release the fully open-source code for gRNAde - our wet-lab validated, generative AI framework for 3D RNA inverse design πβοΈ
I pride myself on open-science & this is probably the most intense release I've done!
I pride myself on open-science & this is probably the most intense release I've done!
December 15, 2025 at 9:26 AM
Excited to release the fully open-source code for gRNAde - our wet-lab validated, generative AI framework for 3D RNA inverse design πβοΈ
I pride myself on open-science & this is probably the most intense release I've done!
I pride myself on open-science & this is probably the most intense release I've done!
Reposted by Chaitanya K. Joshi
Happy to have contributed to and now finally share LeMat-GenBench, a new open benchmark + leaderboard for generative crystalline materials models! βοΈβ¨
It provides standardised metrics for validity, stability, & much more. Already includes results for 12 models!
π Paper: arxiv.org/abs/2512.04562
1/4
It provides standardised metrics for validity, stability, & much more. Already includes results for 12 models!
π Paper: arxiv.org/abs/2512.04562
1/4
LeMat-GenBench: A Unified Evaluation Framework for Crystal Generative Models
Generative machine learning (ML) models hold great promise for accelerating materials discovery through the inverse design of inorganic crystals, enabling an unprecedented exploration of chemical spac...
arxiv.org
December 9, 2025 at 5:05 PM
Happy to have contributed to and now finally share LeMat-GenBench, a new open benchmark + leaderboard for generative crystalline materials models! βοΈβ¨
It provides standardised metrics for validity, stability, & much more. Already includes results for 12 models!
π Paper: arxiv.org/abs/2512.04562
1/4
It provides standardised metrics for validity, stability, & much more. Already includes results for 12 models!
π Paper: arxiv.org/abs/2512.04562
1/4
Reposted by Chaitanya K. Joshi
Thank you to everyone who made the inaugural Virtual Cell Challenge a success.
Over 5,000 participants from 114 countries competed to build AI models that predict cellular responses to genetic perturbations. Today we're announcing the winners and reflecting on what we learned.
Over 5,000 participants from 114 countries competed to build AI models that predict cellular responses to genetic perturbations. Today we're announcing the winners and reflecting on what we learned.
December 7, 2025 at 4:03 AM
Thank you to everyone who made the inaugural Virtual Cell Challenge a success.
Over 5,000 participants from 114 countries competed to build AI models that predict cellular responses to genetic perturbations. Today we're announcing the winners and reflecting on what we learned.
Over 5,000 participants from 114 countries competed to build AI models that predict cellular responses to genetic perturbations. Today we're announcing the winners and reflecting on what we learned.
I think the term βVirtual cellβ will have the same trajectory as βAGIβ or βFoundation modelsβ: Initially opposed by rigorous scientists, while the Bay Area and Demis Hassabis are the only ones comfortable using it
β becoming a mainstream term in academia soon, in few years (Overton window)
β becoming a mainstream term in academia soon, in few years (Overton window)
December 8, 2025 at 7:29 AM
I think the term βVirtual cellβ will have the same trajectory as βAGIβ or βFoundation modelsβ: Initially opposed by rigorous scientists, while the Bay Area and Demis Hassabis are the only ones comfortable using it
β becoming a mainstream term in academia soon, in few years (Overton window)
β becoming a mainstream term in academia soon, in few years (Overton window)
Thankful for the tireless work and motivation from yourself and the Eterna team!
December 4, 2025 at 6:46 AM
Thankful for the tireless work and motivation from yourself and the Eterna team!
Reposted by Chaitanya K. Joshi
An AI researcher interested in biochemistry modeling successfully improved his RNA language model through participation in the Eterna pseudoknot design competition. Congratulations, Chaitanya! π§¬π§ͺ #RNAsky
The polymerase ribozyme results are pretty cool too. π
The polymerase ribozyme results are pretty cool too. π
Introducing gRNAde: our own little "AlphaGo Moment" for RNA design! π§¬π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
December 3, 2025 at 6:16 PM
An AI researcher interested in biochemistry modeling successfully improved his RNA language model through participation in the Eterna pseudoknot design competition. Congratulations, Chaitanya! π§¬π§ͺ #RNAsky
The polymerase ribozyme results are pretty cool too. π
The polymerase ribozyme results are pretty cool too. π
Reposted by Chaitanya K. Joshi
Enumerating possible pseudoknots that a sequence can form with nearest-neighbor models is an NP-hard problem. Even evaluating these structures is challenging, let alone designing them. So itβs great to see data-based models starting to crack the RNA structural design problem! π§¬π§ͺ
Introducing gRNAde: our own little "AlphaGo Moment" for RNA design! π§¬π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
December 3, 2025 at 10:45 PM
Enumerating possible pseudoknots that a sequence can form with nearest-neighbor models is an NP-hard problem. Even evaluating these structures is challenging, let alone designing them. So itβs great to see data-based models starting to crack the RNA structural design problem! π§¬π§ͺ
5/ Want the story behind the science? π°
I wrote a blog post about our "AlphaGo Moment" and what it was like being an AI researcher embedded at the legendary @mrclmb.bsky.social (holding a pipette for the first time!)
Read it on Substack: chaitjo.substack.com/p/alphago-mo...
I wrote a blog post about our "AlphaGo Moment" and what it was like being an AI researcher embedded at the legendary @mrclmb.bsky.social (holding a pipette for the first time!)
Read it on Substack: chaitjo.substack.com/p/alphago-mo...
An AlphaGo moment for RNA design
How our AI system, gRNAde, matched human experts at the complex game of RNA foldingβand why it matters.
chaitjo.substack.com
December 3, 2025 at 6:47 AM
5/ Want the story behind the science? π°
I wrote a blog post about our "AlphaGo Moment" and what it was like being an AI researcher embedded at the legendary @mrclmb.bsky.social (holding a pipette for the first time!)
Read it on Substack: chaitjo.substack.com/p/alphago-mo...
I wrote a blog post about our "AlphaGo Moment" and what it was like being an AI researcher embedded at the legendary @mrclmb.bsky.social (holding a pipette for the first time!)
Read it on Substack: chaitjo.substack.com/p/alphago-mo...
4/ This was a massive team effort bridging AI and biology, from one end of Cambridge to another π€π²
Thanks to Edo Gianni* @edogia.bsky.social, Sam Kwok*, @simonmathis.bsky.social, Pietro LiΓ², and @philholliger.bsky.social for this journey!
π Preprint: tinyurl.com/gRNAde-paper
Thanks to Edo Gianni* @edogia.bsky.social, Sam Kwok*, @simonmathis.bsky.social, Pietro LiΓ², and @philholliger.bsky.social for this journey!
π Preprint: tinyurl.com/gRNAde-paper
Generative inverse design of RNA structure and function with gRNAde
The design of RNA molecules with bespoke three-dimensional structures and functions is a central goal in synthetic biology and biotechnology. However, progress has been limited by the challenges of de...
tinyurl.com
December 3, 2025 at 6:47 AM
4/ This was a massive team effort bridging AI and biology, from one end of Cambridge to another π€π²
Thanks to Edo Gianni* @edogia.bsky.social, Sam Kwok*, @simonmathis.bsky.social, Pietro LiΓ², and @philholliger.bsky.social for this journey!
π Preprint: tinyurl.com/gRNAde-paper
Thanks to Edo Gianni* @edogia.bsky.social, Sam Kwok*, @simonmathis.bsky.social, Pietro LiΓ², and @philholliger.bsky.social for this journey!
π Preprint: tinyurl.com/gRNAde-paper
3/ The Mechanistic Insight π½
What did gRNAde learn? Humans stick close to nature, but gRNAde changes ~70% vs. WT sequence.
And it "sees" invisible 3D constraints that rational methods miss, allowing us to make "generative jumps" to new functional islands in sequence space ποΈ
What did gRNAde learn? Humans stick close to nature, but gRNAde changes ~70% vs. WT sequence.
And it "sees" invisible 3D constraints that rational methods miss, allowing us to make "generative jumps" to new functional islands in sequence space ποΈ
December 3, 2025 at 6:47 AM
3/ The Mechanistic Insight π½
What did gRNAde learn? Humans stick close to nature, but gRNAde changes ~70% vs. WT sequence.
And it "sees" invisible 3D constraints that rational methods miss, allowing us to make "generative jumps" to new functional islands in sequence space ποΈ
What did gRNAde learn? Humans stick close to nature, but gRNAde changes ~70% vs. WT sequence.
And it "sees" invisible 3D constraints that rational methods miss, allowing us to make "generative jumps" to new functional islands in sequence space ποΈ
2/ The Function Challenge βοΈ
But can we move from static shape to dynamic RNA machines?
We show how with a complex RNA Polymerase Ribozyme.
Rational design failed (3% success). gRNAde succeeded (31.5% active), discovering improved variants 15-20 mutations away from nature.
But can we move from static shape to dynamic RNA machines?
We show how with a complex RNA Polymerase Ribozyme.
Rational design failed (3% success). gRNAde succeeded (31.5% active), discovering improved variants 15-20 mutations away from nature.
December 3, 2025 at 6:47 AM
2/ The Function Challenge βοΈ
But can we move from static shape to dynamic RNA machines?
We show how with a complex RNA Polymerase Ribozyme.
Rational design failed (3% success). gRNAde succeeded (31.5% active), discovering improved variants 15-20 mutations away from nature.
But can we move from static shape to dynamic RNA machines?
We show how with a complex RNA Polymerase Ribozyme.
Rational design failed (3% success). gRNAde succeeded (31.5% active), discovering improved variants 15-20 mutations away from nature.
1/ The Structure Challenge π§©
We entered gRNAde into Eterna OpenKnot: a CASP-style blinded, wet-lab competition by @eternagame.org
The result? Parity with the world's best humans, and big gains over Rosetta, RFdiffusion.
We can automate expert intuition for complex RNA folding!
We entered gRNAde into Eterna OpenKnot: a CASP-style blinded, wet-lab competition by @eternagame.org
The result? Parity with the world's best humans, and big gains over Rosetta, RFdiffusion.
We can automate expert intuition for complex RNA folding!
December 3, 2025 at 6:47 AM
1/ The Structure Challenge π§©
We entered gRNAde into Eterna OpenKnot: a CASP-style blinded, wet-lab competition by @eternagame.org
The result? Parity with the world's best humans, and big gains over Rosetta, RFdiffusion.
We can automate expert intuition for complex RNA folding!
We entered gRNAde into Eterna OpenKnot: a CASP-style blinded, wet-lab competition by @eternagame.org
The result? Parity with the world's best humans, and big gains over Rosetta, RFdiffusion.
We can automate expert intuition for complex RNA folding!
Introducing gRNAde: our own little "AlphaGo Moment" for RNA design! π§¬π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
December 3, 2025 at 6:47 AM
Introducing gRNAde: our own little "AlphaGo Moment" for RNA design! π§¬π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π
π: tinyurl.com/gRNAde-paper
Unlike proteins, RNA design has long relied on "wisdom of the crowd" (human experts) or the slow crawl of directed evolution β gRNAde changes that! π§΅π