Tristan Farmer (@001TMF) header
Tristan Farmer (@001TMF) profile picture

Co-founder & CEO Proteus Bio | GitHub: 001TMF | @fdotinc

Joined February 2026
Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)Photo by Tristan Farmer (@001TMF)
Tristan Farmer retweeted
Today we are launching @BioDelivera a programmable drug delivery platform that delivers therapies only to diseased cells enabling cures for every disease. 🎯🧪 Less than 3% of an injected therapy reaches diseased cells. The rest ends up in healthy organs. At @BioDelivera we want to change that, our mission is to make every therapy accessible to patients. Happy to build with @ikeda_trashi this exciting technology! Reach out at [email protected]
36
36
6
289
15,829
Tristan Farmer retweeted
The AI Ring by ORB is ready. How does it work? > The LED shines into the skin, some of that light scatters back and hits the photodiode, and the photodiode converts it into a tiny electrical current. The ring’s electronics turn that current into numbers. Those numbers, recorded over time, are already the raw PPG signal. Each number represents how much light came back at that instant. As your heart beats, blood volume in the tissue rises and falls, changing how much light comes back. That creates a repeating wave in the raw PPG that helps us measure your heart rate and sleep and even stress level. @fdotinc @hthieblot @FurqanR @ruslanjabari
29
20
3
66
1,792
Mythos 5.2 devised a creative (and compute-efficient) solution to folding AAVs using OpenFold3-p2 and ESMFold2! I used to work on these molecules a few years ago @Dyno_Tx so I appreciate the review from @samsinai and the Dyno team. Thank you as well to @MoAlQuraishi and the @open_fold team as well as @ebetica and the @biohub team for developing and releasing these great folding models.
Replying to @claudeai
Gene therapies can be delivered in a protein shell called a capsid. Mythos 5.2 found a new way to predict a whole capsid’s structure. In our tests, it accurately modeled 28 of the 34 capsids it attempted. Biologists could use it to design capsids faster and more cheaply.
4
21
3
212
17,720
Tristan Farmer retweeted
The cloud providers killed our machines 69 times, but we didn't lose a single job. Our coding agent used Portal to run all 54 jobs in our demand forecasting project on spot instances, cutting our cloud bill by 3x ($124 → $43).
10
8
1
37
1,024
Tristan Farmer retweeted
Just noticed there is an antibody specific checkpoint for RFD4-proteina on huggingface since last week. Did I miss an announcement? Any clues of how it was fine-tuned? No info in the preprint huggingface.co/nvidia/RFD4-P…
3
9
1
75
5,062
Tristan Farmer retweeted
We rented the same H100 for 80% less. The catch: the cloud can take it back at any moment, and your job dies with it. So we built Portal. It saves your running Python job and resumes it on any other machine, any cloud, mid-run. No code changes. One-prompt install and forget.
31
29
12
358
33,391
Tristan Farmer retweeted
FinaldDoae November Update ready! finaldose.substack.com/p/202…
1
1
13
5,879
Tristan Farmer retweeted
Great to see teams using the Boltz API at the Ebolathon successfully designing novel molecules! 140 molecules tested, 11 selective hits in Ebola pseudovirus assays. Exciting to see hackathon ideas turn into experimental results! 🧪 muni.bio/research/hits-from-…
2
12
61
4,028
Tristan Farmer retweeted
Introducing Mol-1: Open Frontier Biomolecular Intelligence 🚀🚀🚀 📊 75.0% on FoldBench-AB, near IsoDDE's reported 75.58%!!! 🧬 60k-token assemblies on one single H20, outperforming Anthropic 8/10. 🎯 VHHs binding hit rate as high as 63% with best SPR KD at 43 pM. GitHub: github.com/molence/Mol-1 Try it: lab.molence.com
16
99
13
595
64,638
Tristan Farmer retweeted
Our last challenge in the Anthropic x Adaptyv Protein Design Competition is about immunotherapy, one of the most important ways of treating cancer. Immunotherapy helps the immune system attack cancer cells, but distinguishing them from healthy cells isn't always easy. In some cancers, the difference can be as subtle as a single amino acid in a small protein fragment. That's the case with KRAS, a protein that works like a growth switch in our cells. In many cancers, a mutation leaves the switch stuck on, and the cell keeps multiplying itself and grows into a tumor. In Challenge 5, you need to design a binder that sticks to the mutated piece of KRAS, but not to the healthy version, so that the immune system can find and attack the cancer cells. The deadline to submit protein designs for this challenge is November 8. Then, we'll pick the best designs and test them in our lab.
14
19
11
98
18,326
Tristan Farmer retweeted
Today we are excited to release Artios, a fully open source system for designing functional de novo enzymes. Enzymes present a new frontier for de novo protein design models due to the challenge of coordinating precise chemistry, rather than binding a target. Enzymes are life’s biological machines and by designing them we can truly start to engineer biology. We believe these fundamental capabilities should be built in the open and shared with the community to progress science faster. 🧵 1/6
23
94
11
423
30,850
Tristan Farmer retweeted
Scorpion farmers in shambles. Another job lost to @claudeai 🦂
We're excited to share results from our latest wet lab campaign with @AnthropicAI and @citrini. Claude designed novel proteins to neutralize scorpion toxins and we tested 211 of these protein designs in our automated wet lab. 17 of them showed binding to the toxins in our experiments, with all data now available on @proteinbase. This is a great example of what we're building towards: A world where anyone can run drug discovery campaigns on our automated wet lab.
5
4
56
3,204
Tristan Farmer retweeted
FinalDose is hiring (London)! We're moving full speed ahead with building the most programmable drug that kills cancer cells based on recognizing its DNA/RNA signature. The most awesome payload is useless without the right delivery route.
4
2
13
601
Tristan Farmer retweeted
Good afternoon, San Francisco! Today, we're releasing Test-Time Structure-Space Search, a new OpenDDE inference algorithm for more effective co-folding when designing and selecting antibody drug candidates 🧬. 16% → 78%, that's top-1 success on antibody–antigen targets where OpenDDE usually fails, using Test-Time Structure-Space Search with just five correct contacts. Frozen model. No retraining. No gradients through the network. Test-time compute isn't only for LLMs. Rather than only picking the best of many samples, we search for better structures while the sampler is still denoising. Same OpenDDE, same five contacts, same budget of 25 candidates per target: • re-rank with the contacts: 33% • steer the sampling with them: 78% Re-ranking recovers only about a quarter of the gain. The rest comes from changing the sampling trajectory itself. How it works: • At scheduled denoising steps (about three dozen per sample), we check OpenDDE's predicted clean complex against the contacts. • If they fail, a solver moves the antibody as a rigid body: Refine (nudge locally) → Search (broad pose search, only if still failing) → Refine (polish the best pose). • A clash-aware energy refuses moves that create severe overlaps, and fused GPU kernels make thousands of pose evaluations per call practical. • The corrected structure goes back to the sampler. If the contacts already hold, nothing moves. Results: 805 held-out SAbDab targets where the base model usually fails (median DockQ < 0.23 over five samples). Top-1 success by number of known contacts: 0 (unguided): 16% and 5: 78%. With 5 contacts, mean top-1 DockQ goes 0.14 → 0.43 and the CAPRI-medium rate 11% → 39%. Pocket mode (5 epitope residues, nothing on the antibody side): 17% → 33%, versus 26% for re-ranking. A smaller gain, same pattern. Try it: Put contact pairs or epitope residues in the "constraint" field of your input JSON and run with --use_tfg_guidance true. Checkpoints and output format are unchanged, and inputs without a constraint behave as before. With the optional trunk cache, inputs that differ only in their constraint share one trunk. Examples for Fab, Fv and VHH are in examples/tfg. Fine print: Constraints are simulated from reference structures, so they are correct by construction. This measures what accurate interface knowledge is worth, not robustness to experimental noise, and wrong contacts can push success below unguided. SAbDab only. Success = DockQ ≥ 0.23, 25 candidates per target, 95% bootstrap CIs. Validated on one GPU with 200 diffusion steps. Code: github.com/aurekaresearch/Op… Tech Report: github.com/aurekaresearch/Op…
9
28
6
179
29,116
Tristan Farmer retweeted
You can change it You can influence it You can build your own thing Prompt to hardware blueprint.io
75
42
9
279
18,548
Tristan Farmer retweeted
he is sandboxed for ai safety reasons
Why does Dario Amodei always appear virtually?
143
1,060
53
25,871
566,647
Tristan Farmer retweeted
100 seats. One poker night. SF Tech Week. Founders, scientists & builders around the table. All in on Bio X AI. Link below or DM me. Located @fdotinc.
4
6
2
27
3,987
It's amazing to see BindCraft2 out and I'm even more excited to see what the community can do with it! We've found that Claude excels most when provided with a complex search space to navigate. BindCraft2 expands the search space of molecule design by enabling more modalities and targets, and it's faster. See what you and Claude can do with BindCraft2 in our competition co-sponsored with @adaptyvbio! Sign up here: docs.google.com/forms/d/e/1F… Learn more here: proteinbase.com/competitions…
ʙɪɴᴅᴄʀᴀꜰᴛ2 is out, and we're not waiting for the paper. The full code drops today, free for academic and industry use. We're releasing it early so you can start designing right now, and bring its full power to the current Adaptyv competition. github.com/PacesaLab/BindCra…
1
6
73
7,328
Tristan Farmer retweeted
ʙɪɴᴅᴄʀᴀꜰᴛ2 is out, and we're not waiting for the paper. The full code drops today, free for academic and industry use. We're releasing it early so you can start designing right now, and bring its full power to the current Adaptyv competition. github.com/PacesaLab/BindCra…
32
332
24
1,490
165,182