Today we’re releasing real-world experimental data for over 1000 novel AI-designed proteins on our new platform @proteinbase!
11
104
407
Replies
Proteinbase is the home for protein design data: you can find all protein sequences, their experimental results and design methods in one place. This way you can track performance of protein design models across targets and methods instead of having to search for data scattered
1
1
18
From working with hundreds of protein designers and running thousands of lab validations of AI-designed proteins at @adaptyvbio, we saw a few key points coming up over and over again:
1
1
9
📊 Lack of open, high-quality protein experimental data (including negative data) How we're fixing it: We're releasing thousands of experimental data points generated in the Adaptyv lab, from customers open-sourced their results and from our internal benchmarking campaigns
1
1
12
⚖️ Lack of real-world benchmarks for protein design pipelines How we're fixing it: Proteinbase links every protein to the design method that created it. As data accumulates, you can see how each model performs across different tasks. We're also introducing standardized
1
1
8
🧠 Lack of experimental validation opportunities, which makes it hard to see novel ideas emerge How we're fixing it: We'll organize more protein design competitions that are free to enter, with testing fully funded by Adaptyv or partner organizations.
1
1
9
We have a lot more coming, starting with a new competition that we’ll announce soon. If you’d like to validate your proteins, we’re offering a 20% discount if you open source your results on Proteinbase. Get started here:
1
2
12
@julian_englert @proteinbase Will you include if any of these proteins activate or inhibit particular cellular pathways and if they are specific to it?
0
0
0
@julian_englert @proteinbase Nice work Julian. Can I as an individual researcher after development of these protein design models validate the accuracy of these models myself or I need an expert doctor to validate the accuracy of the model that I develop?
0
0
0
Protein language models can be finetuned to generate many novel structural folds Arjuna Subramanian, Matt Thomson
3
9
79
Thrilled to announce our new preprint, “Protein Hunter: Exploiting Structure Hallucination within Diffusion for Protein Design,” in collaboration with Griffin, @GBhardwaj8 and @sokrypton 🧬Code and notebooks will be released by the end of this week. 🎧Golden- Kpop Demon Hunters
2
40
234
The Baker Lab designs protein on/off switches 1/🧵 Instead of optimizing for how tightly a given protein binds to a target, Adam and team develop a method to control the duration of binding.
3
66
488
Protein design is starting to better explore even the dark & mirror folds not found in nature. DiffTopo explores coarse topology (helices/strands) → guides RFdiffusion + ProteinMPNN/AF2 to backbones. MirrorTopo flips natural folds (L‑aa, right‑handed). It's validated with 5
2
16
104