iiinova — where your digital self keeps evolving

Reposted by Andy Posbe

Sophien Kamoun @kamounlab.bsky.social · 14/09/2026

We just published: Fold on Tight — 10 lessons on AI structure prediction Andy Posbe @adnroide.bsky.social turned our ten lessons on AI structure prediction into a thread. Here it is, one lesson at a time. kamounlab.medium.com/fold-on-tigh...

kamounlab.medium.com

Fold on Tight — 10 lessons on AI structure prediction

Andy Posbe turned our ten lessons on AI structure prediction into a thread. Here it is, one lesson at a time.

01210

Andy Posbe @adnroide.bsky.social · 11/09/2026

No one remembers authorising this firmware update

Five researchers hold freshly printed immune receptor structures while the office Ricoh continues production, despite nobody remembering authorising this feature.

063

Reposted by Andy Posbe

Alex McClelland @alexjmcclelland.bsky.social · 11/09/2026

Very pleased to have contributed to this concise, digestible overview of AI structure prediction! I think this is a great resource for newbies and experienced AlphaFold users alike. Happy Folding!

043

Andy Posbe @adnroide.bsky.social · 10/09/2026

Interface confidence scores ipSAE: github.com/DunbrackLab/IPSAE pDockQ: gitlab.com/ElofssonLab/FoldDock LIS: github.com/flyark/AFM-LIS

000

Andy Posbe @adnroide.bsky.social · 10/09/2026

(3/3) RoseTTAFold All-Atom: github.com/baker-laboratory/RoseTTAFold-All-Atom RF3: github.com/RosettaCommons/foundry ESMFold: github.com/facebookresearch/esm ESMFold2: github.com/Biohub/esm

000

Andy Posbe @adnroide.bsky.social · 10/09/2026

(2/3) Chai-1: github.com/chaidiscovery/chai-lab and lab.chaidiscovery.com Protenix: github.com/bytedance/Protenix and protenix-server.com OpenFold: github.com/aqlaboratory/openfold OpenFold3: github.com/aqlaboratory/openfold-3

200

Andy Posbe @adnroide.bsky.social · 10/09/2026

Structure predictors (1/3) AlphaFold 2 + Multimer: github.com/google-deepmind/alphafold ColabFold: github.com/sokrypton/ColabFold AlphaFold 3: alphafoldserver.com and github.com/google-deepmind/alphafold3 Boltz-2: github.com/jwohlwend/boltz

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Finally, we apologise to colleagues whose excellent work we could not include because of space limitations. Thank you to the community whose discoveries made these ten lessons possible! 🙏

000

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 10 — Develop a Strategy to Stay Up to Date The field moves fast! Follow developers and structural biology communities for new tools and practical tips. Interpret predictions alongside evolutionary and experimental evidence. Full review: zenodo.org/records/2210...

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 9 - Structural Modeling Can Guide Protein Engineering From prediction to design! 🛠️ Interfaces can guide changes that transfer recognition, restore function or create new connections. Combined with binder design, they can help build new immune receptor specificities.

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 8 — When Predictions Deviate from Canonical Models What if the model is telling us something unexpected? Unusual assemblies or repeated failures can hint at different biology. Interpret outliers using existing evidence, then test them. 🔍

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 7 - Capture Touch-and-Go Complexes Brief encounters can have big consequences! AI models can help probe contacts that evade biochemical detection, suggesting how partners meet and which residues to test in the lab. 🧪

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 6 - AI Can Reveal Novel Protein–Protein Interactions From screening entire proteomes to identifying short binding motifs, modelling can uncover new partners and explain how they interact, generating testable hypotheses and guiding experimental discovery.

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 5 - You Can Never Benchmark Enough! A convincing model is a start. How does it compare with what we know? Include known interactors, non-interactors and solved structures as controls. Revisit these controls as tools change. 🧪

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 4 — Structural Similarity Can Reveal Hidden Relationships Proteins with no detectable sequence similarity can share a fold. Comparing predicted structures across families or proteomes can uncover connections that sequence searches miss. 🔍

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 3 — Model Your Proteins with Other Molecules Proteins have company! Lipids, nucleic acids, ions and ligands shape their biology. Including relevant partners can reveal interfaces and improve regions that are poorly modelled in isolation.

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

On tοοls and metrics 🧰 Aside from the lessons, we put together a table of what's currently available and what each tool and metric is actually for. Reproduced below. Links in the reply.

200

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 2 — Sample Multiple Seeds 🎲 How much depends on the seed? Repeat predictions to see which features hold and which vary. Consistency strengthens a hypothesis. Record seeds, model version, weights release and database version so others can reproduce the analysis.

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

Lesson 1 — Don’t Ignore Confidence Metrics 🔍 A model can be uncertain overall yet reliable at the interface that matters. Read local and interface scores alongside global confidence to judge which parts support your biological hypothesis.

100

Andy Posbe @adnroide.bsky.social · 10/09/2026

A pleasure to put this review together with @amiralito.bsky.social @alexjmcclelland.bsky.social @yusugihara.bsky.social @mpcontreras.bsky.social contreras.bsky.social and @kamounlab.bsky.social ! 🌱 Here are the ten lessons 👇

130

Andy Posbe @adnroide.bsky.social · 10/09/2026

Excited to share Fold on Tight! 🎉 We distil recent advances in plant immunity into 10 lessons for using AI structure prediction critically and effectively, in plants and beyond. 📝 zenodo.org/records/2210...

13217

Reposted by Andy Posbe

Jiorgos Kourelis @jiorgoskourelis.bsky.social · 23/01/2026

📢 2 PhD Positions Available in Plant Immune Receptor Bioengineering at Imperial College London 🌱 Start: Oct 2026 📅 Deadline: 10 Feb 2026 🇬🇧 UK/settled status only 💰 Fully funded (48 months) ℹ️ Info: shorturl.at/po61I #PlantImmunity #PlantPathology #SyntheticBiology 🙏 Share with interested candidates!

imperial.ac.uk

PhD Opportunities | Faculty of Natural Sciences | Imperial College London

02131

Reposted by Andy Posbe

Clémence Marchal @clemmar.bsky.social · 03/12/2025

Excited to share our latest work on plant immune receptor biochemistry! @kamounlab.bsky.social @hsuanpai.bsky.social @mpcontreras.bsky.social Jose Salguero Linares @danielluedke.bsky.social @adnroide.bsky.social @jiorgoskourelis.bsky.social www.biorxiv.org/content/10.6...

CC-NLRs exhibit diverse conformations in the resting state.
Figure depicts different NLR resting and activated states. While ZAR1 exists as a monomer bound to its guardees in its resting state, several forms have been observed for the NRC2 helper, including dimers, tetramers, and filaments. While filament-like structures were observed via confocal microscopy for NRC2 upon overexpression in planta, these structures were not detected for its paralog NRC4, suggesting a potentially different resting state configuration for NRC4. Our work suggests that the Pik CC-NLR pair forms a hetero-complex bound to the plasma membrane in its resting state. How effector binding to the integrated domain (ID) of the Pik-1 sensor leads to the activation of the complex remains to be determined.

37639

Reposted by Andy Posbe

Sainsbury Laboratory Cambridge University (SLCU) @slcuplants.bsky.social · 03/11/2025

🎃✨ Pumpkin perfection! We were seriously impressed by the beautifully carved and brilliantly lit pumpkins from @thesainsburylab.bsky.social🕯️ We also got spiritedly competitive with our annual team pumpkin carving competition 😄🎃 Who knew plant scientists were such talented artists? 🪚🎃🌿

Collage of 13 pumpkins that have been carved in various designs and painted for the annual SLCU Pumpkin Carving Competition

1152

Andy Posbe @adnroide.bsky.social · 30/10/2025

Spooky season has arrived! 👻 We held our annual Pumpkin Carving Competition at @thesainsburylab.bsky.social —vote for your favourite carve! 🎃🕯️

8198

Reposted by Andy Posbe

Clémence Marchal @clemmar.bsky.social · 04/09/2025

Super excited and honoured to have been awarded an ERC Starting Grant for CooPAIRation! How do plant immune receptors work together to trigger an immune response? Can we modulate immune responses for disease resistance? We will tackle these questions in CooPAIRation Positions to be announced soon!

188317

Reposted by Andy Posbe

Josh Bennett @joshwbennett.bsky.social · 14/07/2025

Excited to attend my 1st ISMPMI conference, and present our work on a paired wheat CNL/MLKL receptor mechanism! 🌾 Come chat to me at my poster on Tuesday, P-050! 📜 If you miss my poster session, I will be talking in: Crop resistance genetics and genomics on Wednesday - 16:25-16:35!

03114

Reposted by Andy Posbe

Morgan Carter @morgancarterphd.bsky.social · 13/07/2025

Incredibly inspiring talk about the discovery of the more than 200 immune systems in bacteria by Rotem Sorek. From computational discovery in 2018 through functional and structural analyses, even linking bacterial immune systems to that of eukaryotes. #2025ISMPMI

A large conference stage with a man presenting with a power point slide showing a scientific figure about bacterial immunity

25213

Reposted by Andy Posbe

Current Biology @currentbiology.bsky.social · 19/11/2024

You asked for it...

401832324

Reposted by Andy Posbe

Angus Bucknell @ahbucknell.com · 26/10/2023

Great time at the TSL student/postdoc pumpkin carving social today! 🎃 📸 by Matt Wengler

25 carved pumpkins in a line lit with tea lights. The reflection of the pumpkins can be seen in the water of a pond in front of them.

0237