muni

18 posts

muni banner
muni

muni

@muni_bio

An agentic platform for scientific discovery, connecting computational exploration with experimental validation

San Francisco Katılım Aralık 2025
3 Takip Edilen181 Takipçiler
muni retweetledi
Tudor-Stefan Cotet
Tudor-Stefan Cotet@CotetTudor·
It was fun working with @katyenko at @muni_bio on this hackathon over the past few weeks. She did an incredible job organizing it and we had some really cool results to talk about. Out today + check out the analysis blog post!!
Adaptyv Bio@adaptyvbio

What happens when you let frontier LLMs design proteins, and then synthesize and test them in a wet lab? We ran a protein design competition with @muni_bio where AI agents competed against humans to design molecules that bind TREM2, a key receptor linked to Alzheimer’s. Results: GPT 5.2 and Grok 4.1 both placed in the top 5, with molecules showing strong binding to TREM2 when tested in our lab.

English
1
2
12
650
muni retweetledi
Martin Pacesa
Martin Pacesa@MartinPacesa·
The 2 nM BindCraft design is pretty cool
Martin Pacesa tweet media
Adaptyv Bio@adaptyvbio

What happens when you let frontier LLMs design proteins, and then synthesize and test them in a wet lab? We ran a protein design competition with @muni_bio where AI agents competed against humans to design molecules that bind TREM2, a key receptor linked to Alzheimer’s. Results: GPT 5.2 and Grok 4.1 both placed in the top 5, with molecules showing strong binding to TREM2 when tested in our lab.

English
3
16
101
8.5K
muni retweetledi
Kat
Kat@katyenko·
The binders have bound! A few months ago, 9 human teams and 6 autonomous AI agents spent a single day designing protein binders against TREM2 on @muni_bio, a target implicated in Alzheimer’s Disease. 141 designs were submitted, 100 were synthesized and tested by @adaptyvbio, and 37 bound. And surprisingly, AI agents essentially matched human teams on hit rate. These aren’t benchmark scores or simulated results, but real proteins designed in one day in SF and validated experimentally during the first large-scale test of muni, where teams ran 260 GPU jobs and generated a total of 4,176 binders. We wrote about what we learned from the results, how well ipSAE worked as a scoring function, and how this hackathon reshaped what we’re building: muni.bio/research/agent…
English
4
15
73
4.5K
muni retweetledi
Rowan
Rowan@RowanSci·
We had a great time at the AI x Med Chem TBXT hackathon this weekend—thanks for everyone who came to the event (some from quite far away)! Next step: synthesis + testing.
Kat@katyenko

20 pizzas and 7 hours later, we finished the first leg of the AI x Med Chem Hackathon in Boston, where teams competed to submit compounds for TBXT. This is the first hackathon of its kind, moving from small molecule compound generation to experimental assays. Thank you to everyone who joined us this weekend! Huge thanks to our judges, partners and sponsors: @RowanSci, @onepot_ai, @anyscalecompute, @pillar_vc and the @ChordomaFDN. Compound synthesis is underway and will be tested soon! Stay tuned for updates.

English
0
3
13
1K
muni retweetledi
Adaptyv Bio
Adaptyv Bio@adaptyvbio·
What happens when you let frontier LLMs design proteins, and then synthesize and test them in a wet lab? We ran a protein design competition with @muni_bio where AI agents competed against humans to design molecules that bind TREM2, a key receptor linked to Alzheimer’s. Results: GPT 5.2 and Grok 4.1 both placed in the top 5, with molecules showing strong binding to TREM2 when tested in our lab.
Adaptyv Bio tweet media
English
4
15
76
23.9K
muni retweetledi
Kat
Kat@katyenko·
20 pizzas and 7 hours later, we finished the first leg of the AI x Med Chem Hackathon in Boston, where teams competed to submit compounds for TBXT. This is the first hackathon of its kind, moving from small molecule compound generation to experimental assays. Thank you to everyone who joined us this weekend! Huge thanks to our judges, partners and sponsors: @RowanSci, @onepot_ai, @anyscalecompute, @pillar_vc and the @ChordomaFDN. Compound synthesis is underway and will be tested soon! Stay tuned for updates.
Kat tweet mediaKat tweet mediaKat tweet mediaKat tweet media
English
2
6
14
1.8K
muni retweetledi
Kat
Kat@katyenko·
Traditional FEP is slow and prohibitively expensive, but @RowanSci FEP changes this. In this article, we show how rapid FEP can be placed earlier in the loop, connect to automated analogue enumeration in @muni_bio, and bring molecules straight into synthesis with @onepot_ai. muni.bio/research/fast-…
English
1
8
34
3K
muni
muni@muni_bio·
LAN parties but for science. Come join.
English
3
2
9
593
muni
muni@muni_bio·
@AllThingsApx This would be really fun to put together. want to collab? We plan on doing a few agent runs at our next event. If you're in Boston, you're welcome to come check it out. luma.com/n9hheb8j
English
1
0
6
68
Kyle Tretina, Ph.D.
Kyle Tretina, Ph.D.@AllThingsApx·
It's pretty cool to see orgs like @muni_bio experiment with protein design loops in the context of autoresearch. I wonder what it would look like for someone to host a CASP-like competition where only agents (or agent swarms) could compete. Maybe include a few humans for baselines.
Kyle Tretina, Ph.D. tweet media
English
7
4
32
2.2K
muni retweetledi
Derek Alia
Derek Alia@derekalia·
This is the most ambitious agentic system I’ve built. Shoutout to @Cloudflare for dynamic workers and @Anyscale for cluster management. In Muni, every node is a code container. Agents can sift through a virtual file system, write code against the data they need, call models, run jobs, and render the result back onto the canvas. That flexibility matters because the landscape of AI x Bio is moving so quickly, and it’s not obvious what the future scientific workflows will look like. Our goal is to build primitives that adapt as the models, data, and workflows change, so scientists can focus on the questions, not the tooling. muni.bio/research/muni-…
English
1
4
46
275.2K
muni
muni@muni_bio·
bioArena has renamed to muni! The mission is the same, but the new name comes from 'muni-muni' in tagalog (to reflect/ponder/contemplate) and from SF's Muni, and feels closer to both what we’re building and where we’re building it. More soon from @muni_bio.
Kat@katyenko

TBXT (Brachyury) is a key driver of chordoma, a rare bone cancer, and a notoriously difficult target for small molecule discovery. On May 9th, we’re hosting a design-to-assay hackathon in Boston where teams will use the latest scientific models to design small molecules for TBXT. With support from @RowanSci, top compounds will be synthesized by @onepot_ai and move into real binding assays run by the @ChordomaFDN, with compounds that meet the published assay criteria eligible for up to $250K in TBXT Challenge prizes. If AI is going to have real value in drug discovery, these models should be tested against real targets. Hosted by @pillar_vc: luma.com/n9hheb8j

English
0
0
5
264
muni retweetledi
Kat
Kat@katyenko·
This was a huge success. For the first time, we watched scientists and engineers iterate on a real challenge and present solutions in under 2 hours. It’s rare to see scientists at a hackathon, and this is clearly the beginning of something very special. An obvious missing piece here is real-world experimental validation, and this is exactly what we’re adding to the next event with @adaptyvbio. Stay tuned! Thank you to everyone who joined the first agentic bio hackathon - and to @TrueVentures for generously hosting, @RowanSci and @OpenRouter for co-sponsoring, and our fantastic judges @AriWagen, @geneticdrifts, @JaredXia , and Bennett.
Kat tweet mediaKat tweet mediaKat tweet mediaKat tweet media
Kat@katyenko

For the AI-curious scientists and bio-curious engineers: we’re hosting an agentic bio hackathon as a practical way to test AI tools on real scientific problems. Hosted by @trueventures and co-sponsored by @RowanSci. Come join us on Thursday, Jan. 29th luma.com/0l5w97j0

English
2
10
47
14.7K
muni retweetledi
Kat
Kat@katyenko·
@bioArena_ is adding more models to the rotation. These additions are based on what currently supports tool calling for agents:   1. anthropic/claude-sonnet-4.5   2. google/gemini-2.5-pro   3. google/gemini-3-pro-preview   4. openai/gpt-5   5. openai/gpt-5.2   6. deepseek/deepseek-v3.2   7. x-ai/grok-4.1-fast   8. z-ai/glm-4.6   9. minimax/minimax-m2   10. mistralai/mistral-large Thanks to @openrouter for making this easy. 🧵
English
1
1
6
510
muni retweetledi
Kat
Kat@katyenko·
Payouts are live, which means you can now get rewarded for your scientific domain expertise. The reward is shown at the top right of every battle. Upon winning, USDC gets dropped directly into your account. Win battles - your gel is taking time to run anyway. 🧵
English
2
4
9
768

Keşfet

@katyenko @adaptyvbio @RowanSci @onepot_ai @anyscalecompute @pillar_vc @ChordomaFDN @AllThingsApx