Went to check on how #DeepMind did at #CASP15 and discovered that #AlphaFold did not even enter. Instead, most of the participants used AlphaFold 2 or some variant. So DeepMind has revolutionized protein folding prediction. With all the hype around #AI, why do real gains like this get ignored in favor of toys?

On Wednesday Jan 11, at 17:00 (Swedish Time) we will have the first CASP zoom meeting. All information can be found at casp.bioinfo.se. Here is a short recap:

Mailinglist: casp-zoom@googlegroups.com

Zoom: https://stockholmuniversity.zoom.us/j/61626785695

 The first meeting will be dedicated to discussing how to proceed with these meetings.

#casp15

My take on the last year of Protein structure prediction up and including #CASP15 comments are welcome.
[2212.07702] Protein Structure Prediction until CASP15 https://arxiv.org/abs/2212.07702

My take on the last year of Protein structure prediction up and including #CASP15 comments are welcome.
[2212.07702] Protein Structure Prediction until CASP15 https://arxiv.org/abs/2212.07702

#CASP15 in Nature https://www.nature.com/articles/d41586-022-04438-1

#CASP15 in Nature is out now https://www.nature.com/articles/d41586-022-04438-1

RT @dxbarradas@twitter.com

Last day of #CASP15 - discussions about what is an interface ? Very interesting 😃

🐦🔗: https://twitter.com/dxbarradas/status/1602572525480935424

Slides of the different #CASP15 presentations available for download.
Slide below shows how #CASP14 (black line) was truly the renaissance moment of protein folding

RT @sokrypton@twitter.com

#CASP15 slides from days 1-3 released!
https://predictioncenter.org/casp15/doc/presentations/

🐦🔗: https://twitter.com/sokrypton/status/1602308662260908032

More notes from the #CASP15 competition. Other resources here: https://www.aditiashenoy.com/posts/casp15/day2/

RT @AlbertVilella@twitter.com

Some initial results from the #CASP15 competition and we see that #Alphafold2 has become a fertile ground for experimentation by several research groups around the world. Facebook's ESM protein language models (pLMs) are the top non-MSA based methods.

🐦🔗: https://twitter.com/AlbertVilella/status/1601909663309901824

Another take-away from #CASP15 is that people are now starting to consider large-scale computational Protein Protein Interaction analyses (PPIs). These have been done experimentally for more than 30 years in a high-throughput manner, e.g. for yeast. https://www.aditiashenoy.com/posts/casp15/day2/

Some of the challenges of ligand prediction within #CASP at #CASP15 . Large flexible ligands still a challenge. Need for additional metrics/challenges?

Continuing with #CASP15, now on the topic of ligands and their prediction on existing protein structures.

#casp15 ema methods are complicated

RT @bjornwallner@twitter.com

⁦@gabriielstuder@twitter.com⁩ QA assessor in #CASP15 reminds us about the length dependence of TMscore. Problematic for long sequences.

🐦🔗: https://twitter.com/bjornwallner/status/1602549513314570241

#casp15 target matters

#casp15 ema evaluation

RT @aditiashenoy@twitter.com

Takeaways from Day 3 at #CASP15
https://www.aditiashenoy.com/posts/casp15/day3/

🐦🔗: https://twitter.com/aditiashenoy/status/1602419271337910272

#casp15

This is from discussion yesterday and will be discussed further today give me your thoughts

My suggestion is that we schedule monthly meetings at the last Monday in the month at 17:00 (5pm) Swedish time. This is (normally), 16:00, UK time, 11:00 am in New York, 8:00 am San Francisco, 22:00 in Beijing, I think. 

Lots of content about #casp15 on the incipient #proteinengineering mastodon! I missed this exciting preprint using Bakerlab's RF Diffusion led by Susana Vazquez Torres to bind small helical peptides. Very hard design target with pre-DL Rosetta methods. https://www.biorxiv.org/content/10.1101/2022.12.10.519862v2

Lots of content about #casp15 on the incipient #proteinengineering mastodon! I missed this exciting preprint using Bakerlab's RF Diffusion led by Susana Vazquez Torres to bind small helical peptides. Very hard design target with pre-DL Rosetta methods. https://www.biorxiv.org/content/10.1101/2022.12.10.519862v2