Following the lab’s bird theme, we’re excited to present ALBATROSS; A deep-Learning Based Approach for predicTing pRoperties Of diSordered proteinS).

Very nice! I confirm that approach works - we're working on sth similar in AlphaPulldown too. Coming soon! (github.com/KosinskiLab/Al…)

Super fast and versatile toolbox for both single-particle #cryoEM and in-situ #teamtomo cryo-ET! Our new pyTME library consolidates particle picking/structure fitting/template matching in a single framework. Preprint: biorxiv.org/content/10.110… Code &howto: github.com/KosinskiLab/py…

Congratulations, Jan Kosinski, Julia Mahamid, and Georg Zeller! The three EMBL scientists have received prestigious ERC Synergy Grants to enable ambitious collaborative projects that aim to study complex molecular and cellular interactions in context. embl.org/news/science/e…

A new version of AlphaPulldown by DingquanYu. Includes experimental support for multimeric templates in AlphaFold by Dmitry Molodenskiy and interface to crosslink-based modeling AlphaLink2 of Rappsilber laboratory github.com/KosinskiLab/Al…

Work is in full swing on AlphaPulldown by DingquanYu and Dmitry Molodenskiy, but with so many users, thousands of downloads, and 34 open issues, incl. many exciting feature requests, help would be welcome! If you want to contribute, feel free to fork and submit a pull request!

I’m grateful for the opportunity of contributing to this project. Very glad to see experimental evidence and structural predictions work side by side to solve the molecular puzzles. Many thanks to EMBL, CSSB, and BNITM for the great support.

The final version of the PyTME paper published in SoftwareX Journal! PyTME enables fast CPU and GPU-based template matching in #cryoET and fitting in #cryoEM. More in the thread with plenty of screenshots👇 sciencedirect.com/science/articl…

We are recruiting two PhD students 🎓 this round! Dry lab 💻: AI-based structural modeling method development and application. Wet lab 🧪: In-cell cryoET and/or crosslinking mass spectrometry of viral infection. Apply ✍️ by 11 March! ⌛️

I’m greatly thankful to Jan Kosinski and Mohammed AlQuraishi for this opportunity. The openfold team has provided me with invaluable help and inspirations. Excited for the upcoming developments in OpenFold!

Confidence metrics (e.g. pLDDT, pAE) can proxy designability with promising experimental success rates and also reveal alternative states (structures shown are generated by AFCluster).

Here is a AlphaFold/ColabFold tutorial we have recently presented - from basic to more advanced usage, enforcing alternative conformations, modeling interactions, and analysis with ChimeraX and PICKLUSTER. Maybe you find it helpful 🎓 ! Commenting enabled in the doc so feel free

ScaleFold: Reducing AlphaFold Initial Training Time to 10 Hours arxiv.org/abs/2404.11068

AlphaFold3 is insane. MutS protein (green) recognizes base mismatches in DNA (gray), after which it slides off to find MutL matchmaker protein (blue), which positions and activates MutH nuclease (pink) to nick a hemimethylated GATC sequence (red) to activate removal and repair of

OpenFold paper is now out in Nature Methods. Timely. I’ll say more about AF3 at the end of this thread, but first I wanted to highlight what’s new since our initial release. (1/9)

bioRxiv Bioinfo Thanks Alexandre Bonvin - also on Bluesky for testing this and citing and using AlphaPulldown! In AlphaPulldown you can also limit to one model by selecting one model name and using it as the input for the --model_names flag. We can try to emphasize this in the new manual we’re working on, but maybe

Two 🔥 computational postdoc positions 🔥are available in our lab, funded by European Research Council (ERC) ! embl.org/jobs/position/… and embl.org/jobs/position/…. Join us and unleash the power of these just-delivered beasts 👇And please repost to help us!

We have released version 2 of AlphaPulldown - a general pipeline for high-throughput structural modeling! Software: github.com/KosinskiLab/Al… Preprint: biorxiv.org/content/10.110…