So, OpenAI, whose entire business model relies on some overhyped dudebros using vast amounts of stolen data and stolen labour, is releasing a specialised GPT for the life sciences. And naming it, apparently without any sense of irony or self-awareness. Rosalind.
I haven't been following the RL literature closely but I find it hard to believe this hasn't been tried before. I want it to work; anyone tried it?
This is really bad. The ERC is the lifeline of many labs. The letter from the ERC says “the only way is to reduce applications” is the wrong direction!
Use AI to assist review. It’s not perfect, but it allows more proposals to be evaluated, and it should support, not replace, human reviewers.
🤩
My daughter was invited on stage to help Neil deGrasse Tyson @neildegrassetyson.com post a Tweet!
Run an MD simulation of any protein in the AlphaFold Protein Structure Database using AF-CALVADOS
Thanks to @sobuelow.bsky.social AF-CALVADOS is now on Colab
colab.research.google.com/github/KULL-...
DROP-CARs: Engineering Reversible, Drug-Controlled CAR T Cell Activity with a Clinically Approved Small Molecule | Wendell Lim @ucsanfrancisco.bsky.social Adam Stevens @upenn.edu via @theaacr.bsky.social url: aacrjournals.org/cancerres/ar...
Really fun and challenging (!) discussion with @cdbahl.com on BindCraft, the state of binder design, the oustanding challenges and perhaps also some misconceptions in the field 💫
www.healthtech.com/the-chain/ma...
initial model - class 3d - post-processing
it’s been a while. Since I have no local workstation for data processing in my new role yet, I’ve been playing with AWS.
In my downtime, I asked Claude Code to follow a RELION tutorial from scratch. With a few fixes, I got a 3.22 Å result.
Not bad for AI handling routine tasks! :P
We just passed an exciting milestone: our one thousandth member just joined @bpdmc.org!
Thank you so much to everyone who has made this journey possible, and we're excited to welcome many more wonderful folks to our nerdy protein club!
For more info, check out bpdmc.org
an xray density map depicting blue meshed density where an atomic model has been built.
Some people cleave tags religiously. This person didnt and made me solve their perfectly structured Histag. Probably also helped in packing. Might have to do experiments without it to show its not an artifact though.
Team cleave or not cleave?
2D classes of a 58 kDa particle collected on a 100 kev Tundra electron microscope
2D classes of the 58 kDa sample collected on the 100 keV Tundra TEM with a Falcon C detector at @nusbf.bsky.social @newcastleuni.bsky.social. Still much work to do to optimise the sample & data collection but its cool to see what a 100 keV TEM is capable of. #cryoEM @recon4imd.bsky.social
Dear colleagues, With deep sadness I share the news that Andreas Engel passed away last Wednesday, at the age of 82, after a courageous fight against an aggressive caner. Only half a year ago, he was out hiking in the Italian mountains with his wife Barbara, walking 160 km in a single week. Andreas trained as a physicist in Bern, Switzerland, before heading to John Hopkins University in Baltimore for his postdoctoral work. He then led a research group at the Biozentrum in Basel, and in 1985 took on the role of a group leader in research and development at Ilford Ltd. in Fribourg, Switzerland. In 1987, he joined forces with his colleague Ueli Aebi to establish the Maurice E. Müller Institute at the Biozentrum of the University of Basel. Together, Ueli and Andreas served as professors for structural biology there for almost a quarter century, leaving a lasting mark on the field of structural biology as we know it today. Andreas Engel was a true pioneer. He was among the first to apply atomic force microscopy (AFM) with the tip operating under water, and one of the earliest researchers to perform 3D reconstructions using scanning transmission electron microscopy, achieving remarkable high-resolution maps of porin 2D crystals by STEM. He became a leading expert in membrane protein 2D crystallization, and his group played a key role in the determination of the first aquaporin structure, in collaboration with Peter Agre and Yoshinori Fujiyoshi. As he approached retirement from Basel, Andreas founded the Center for Cellular Imaging and NanoAnalytics (C-CINA) of the University of Basel, which in 2009 operated among others one of the very early Titan Krios instruments. From 2008 also held a professorship at Case Western Reserve University, Cleveland, followed by a position at TU Delft in the Netherlands in 2013. In 2020, he was among the founding members of CryoWrite, a company in Basel that he led as CEO until recently. His contributions to science, to the tools of str…
Photo of Andreas Engel
Sad news on the passing of Prof. Andreas Engel, a pioneer in the field of cryo-EM (shared with the 3DEM community by Henning Stahlberg).
Andreas' kindness, warmth, and enthusiasm for science and his colleagues will be missed.
The Critical Assessment of Structure Prediction (CASP) experiment is calling for prediction targets: Immune Complexes, Organic Ligand-Protein Complexes, Nucleic Acids and Complexes, Conformational Ensembles, Difficult Protein Structures and Complexes. Rule of Thumb: If AlphaFold3 can generate a high-quality model, it is likely not a CASP-grade challenge. If it struggles, we want it.
Is #AI hitting a plateau in structure prediction? Help us find out at CASP17! 🧪🧬
Calling for Targets: Immune Complexes, protein - ligand complexes, RNA/DNA, conformational ensembles, membrane proteins, viral origins, and large complexes.
The Rule of Thumb: If AF3 can’t model it, we want it.
On this 1st April I'm happy to release a technical report from our startup MirageBio, describing a new protein structure model with state-of-the-art performance.
jgreener64.github.io/posts/techni...
Oh thank god someone finally wrote this
AI is powerful, but hype gives it pseudoscience vibes.
When claims are unfalsifiable, detached from reality, or driven by tools looking for problems, we risk undermining real progress.
Overselling AI does not accelerate science: It erode trust.
stevensalzberg.substack.com/p/ai-is-star...
ROCKET 🚀 inference-time optimization of AlphaFold to fit structural data is published! rdcu.be/fa9YH
Since our preprint, we’ve pushed it to regimes where other methods break: low resolution, weak signal, real experimental edge cases. Here’s what we learned: 1/15
Protein language models are bad at mutational effect prediction
blog.genesmindsmachines.com/p/protein-la...
It is probably only useful to me, but just in case I have put it on github - caveat emptor, I have not tested it extensively, but it seems to give sensible results. I just like to have an annotated sequence handy while building, and this scratches that itch. (3/3) github.com/olibclarke/p...
An image of the Padrãos does Descobrimentos (Monument to the Discoveries) sculpture in front of purple and orange lit clouds during sunset. The portion of the sculpture shown is primarily white colored and shows three men dressed in clothes from the late 15th century with the man in the front holding a model of a sailing ship. The text on the image says, “European RosettaCon, Lisbon, Portugal, Crossing Boundaries with Protein Design, Registrations are open, Early bird until 31st of May, 28th to 30th of October, 2026, NOVA University Lisbon, Campolide Campus, Rectory.” It has the Rosetta Commons logo at the bottom of the image in the middle.
European RosettaCon invites researchers in protein design, biomolecular dynamics, advanced microscopy, and structural determination methods to “cross boundaries in protein design” together. Join us to reach beyond traditional frontiers and into interdisciplinary approaches. europeanrosettacon.org
Our paper with @sokrypton.org using AlphaFold2 to predict small-molecule binding sites in proteins is now out in Nature Methods! 🧵
rdcu.be/e7SnX
www.nature.com/articles/s41...
Excellent work
New paper showing that much of the apparent success of protein language models in predicting mutational effects is a mirage: These models mostly memorize sites. 1/
www.biorxiv.org/content/10.6...
We built DNA scaffold crystals that can be reprogrammed to host different guest proteins for X-ray structure determination — no new crystallization screen required. This animation shows a guest protein (glowing) seeking its designed binding site inside the crystal lattice. shorturl.at/DANdc
Happy to share our new preprint on the mechanism of human tRNA 3' CCA maturation! This project was spearheaded by Bernhard Kuhle in my group, with contributions from many others and a great collaboration with the group of Peter Rehling (UMG). See highlights below!
www.biorxiv.org/content/10.6...
Low resolution cryo-EM map that looks like Cuban's skull
When you still haven't solved the orientation bias of your protein sample #CryoEM
We have started a project trying to predic the interactions/structures of all yeast protein pairs using an AlphaFold pooling approach. We are making the current dataset open and we welcome collaborations.
www.evocellnet.com/2026/03/mapp...
Almost forgot to brag about the absolutely amazing office the UZH Insitute of Pharmacology has provided for me ♥️ Protein structural art by Francisco Enguita (www.behance.net/fenguita)
