AlphaFold Unofficial
banner
alphafold.bsky.social
AlphaFold Unofficial
@alphafold.bsky.social
Unofficial account exploring the intersection of biology, molecules, science, AI and protein folding with AlphaFold.
Reposted by AlphaFold Unofficial
CB features: ✅ Diverse protein types (receptors, enzymes, synthetic binders) ✅ Monomers & oligomers ✅ Natural or synthetic (no MSA needed) ✅ Input: 2 structures (PDB or AlphaFold) ✅ Fast: ~3k variants in <1 min (RTX 4090) ✅ Scores single & multi-mutants

Code & user-friendly interface here:
GitHub - alicetinglab/ConformationalBiasing: Code for designing biased protein states
Code for designing biased protein states. Contribute to alicetinglab/ConformationalBiasing development by creating an account on GitHub.
github.com
January 8, 2026 at 7:12 PM
Reposted by AlphaFold Unofficial
💻 Using the #fragment #screening analysis from our #FRAGSYS pipeline, combined with high quality #Google #Deepmind's #AlphaFold #Multimer (ColabFold) models of the #SACK1G dimer, we identified and #prioritised 13 #residues in a binding site of interest as likely #functionally #relevant.
January 9, 2026 at 10:41 AM
Reposted by AlphaFold Unofficial
Things like AlphaFold were great back when we had a scientific infrastructure. Without the PDB and CASP, how far would it have got? It’s a major issue when gigawatts of power are being diverted to train universal function approximators to come up with more believable lies.
January 11, 2026 at 4:33 AM
Reposted by AlphaFold Unofficial
We call this approach Conformational Attention Analysis Tool (CAAT). Applying CAAT to fold switcher KaiB revealed 4 important sites, 3 of which had been identified previously through hundreds of AlphaFold runs. CAAT requires just 2 AlphaFold passes. It suggested a new double mutant.
January 5, 2026 at 3:28 PM
Reposted by AlphaFold Unofficial
Meiner Meinung nach beginnt die Geschichte nicht mit Männern, die sich in Dartmouth treffen, sondern mit einer Frau.

Schauen wir uns mal diese (subjektive, stark reduzierte) Timeline an.
January 8, 2026 at 3:24 PM
Reposted by AlphaFold Unofficial
Der nächste Durchbruch kam im Jahr 2014 mit Ian Goodfellows Paper zu GANs (Generative Adversarial Networks), Modellen zur Generierung von Bildern. Danach kam eine Innovation nach der anderen: KI schlägt Mensch in AlphaGo, AlphaFold sagt Proteinstrukturen vorher, vor gut 3 Jahren …
January 8, 2026 at 3:33 PM
Reposted by AlphaFold Unofficial
You're getting mad at straw men. I follow tons of anti-GenAI people and I haven't seen any of them complain about AI being used for things like Alphafold. People hate GenAI because it stole the works of thousands of artists/writers and is being used to drive those same creatives out of work
December 21, 2025 at 7:29 PM
Reposted by AlphaFold Unofficial
You ignore the progress that is being made in science by non scientists with AI. Alphafold, and material sciences use AI with a lot of success. It's not an Oracle that replaces your work. It can be a great enhancement, and produce slop just like most humans ,😄. Yes its here and will change stuff.
December 21, 2025 at 6:11 PM
Reposted by AlphaFold Unofficial
Can we please please please start articles about AI with a disambiguation about what kind exactly we are talking about?
AlphaFold is very much different from generative AI, both in the underlying technology and in domain it is used.
December 21, 2025 at 3:32 AM
Reposted by AlphaFold Unofficial
The ONLY ‘AI’ I will allow are specialist models, like Alphafold. Things that are actively working to improve lives. Not taking jobs or profiting off of others’ work. Not GenAI or LLMs.
December 19, 2025 at 10:28 PM
Reposted by AlphaFold Unofficial
AlphaFold was of course a version of that too, even if it couldn't score well here. But it wasn't general. These capabilities appear to be general intelligence, e.g. according to Francois Chollet's well considered framing here arcprize.org/arc-agi
ARC Prize - What is ARC-AGI?
The only AI benchmark that measures AGI progress.
arcprize.org
December 19, 2025 at 9:02 PM