Daniel Mattox

Excited to unveil Boltz-2, our new model capable not only of predicting structures but also binding affinities! Boltz-2 is the first AI model to approach the performance of FEP simulations while being more than 1000x faster! All open-sourced under MIT license! A thread… 🤗🚀

KAAPO KRAKEN!!! 🦑

@swcanner Very cool work! Some parallels to some of my thesis work, I can definitely commiserate on pulling out lectin interactions from glycan-containing PDB files journals.plos.org/ploscompbiol/a…

Hello hello, I’m excited to share our most recent work on carbohydrate-protein interactions. CArbohydrate-Protein Site IdentiFier (CAPSIF) is a carbohydrate-binding site prediction model for carbohydrate binding sites on a protein. biorxiv.org/content/10.110…

@AdaptiveBiotech does antibodies too! Check out this freshly published collaborative project: Bivalent intra-spike binding provides durability against emergent Omicron lineages: Results from a global consortium cell.com/cell-reports/f…

Statement WWWWWWWin!!

C-mannosylation of tryptophans is critical for receptor function and trafficking in metazoans, viruses, and malaria parasites. I am thrilled to finally present the structural basis of this mechanistic puzzle! Thank you @ethangbaus, Kosslab, @reymondgroup! nature.com/articles/s4158…

A pleasure and a privilege to be assessor again at #CASP15. A mix of spectacular success and the occasional intriguing failure. Looking forward to hearing from the predictors and discussing future directions. Thanks @AJSimpkin @rmk_ccp4 @shah_comp526 @DavidLMurphy3 Filo and Luc.

Our preprint on OpenFold, our trainable reproduction of AlphaFold2, is finally up (biorxiv.org/cgi/content/sh…)! Since we open-sourced parameters in June, we've trained the model to high accuracy more than 50 times, on a variety of datasets. Here's what we learned (a lot) -> (1/19)

BREAKING NEWS: The Royal Swedish Academy of Sciences has decided to award the 2022 #NobelPrize in Chemistry to Carolyn R. Bertozzi, Morten Meldal and K. Barry Sharpless “for the development of click chemistry and bioorthogonal chemistry.”

GOODBYE BASEBALL. HELLO POSTSEASON.

Preprint from Philip Bradley where he creates a version of AlphaFold to model TCR:peptide-MHC interactions. Benchmark is far from perfect, but the paper shows that deep learning-based structural modelling is a possible strategy to predict TCR specificity. doi.org/10.1101/2022.0…

1/6 Can #AlphaFold learn from + & - labeled data to predict peptide binding better? w/ Phil Bradley @JustasDauparas @minkbaek @Mohamad_Abedi David Baker we present joint structure prediction-classification fine-tuning of #AF2 to classify pMHC binding.@UWproteindesign @fredhutch

@pjie2 @surt_lab @pmoyniha I really appreciate this point, but I’m curious if you think the homology filtering wouldn’t account for this. I would say that 70% seems like a pretty lenient filter and I would be interested to see if the results hold with more conservative filtering.

This is conceptually insane. nature.com/articles/s4146…

May 5 is a national day of awareness for missing and murdered Indigenous women. For more than a year, reporters across @usatoday network have worked together to examine factors contributing to the crisis and how laws have largely failed to address it.#MMIW usatoday.com/in-depth/news/…

Hi Joe President, I have a question

Studying protein without its glycans is like studying a bird without a single feather; even the best-designed and performed study will never see it flying… Our review on developments and perspectives in high-hhroughput glycomics just got published. academic.oup.com/glycob/advance…

A view of today’s afternoon hail storm from west Seattle @KSeattleWeather

@AITA_online He’s only 12.5% boxer but my good boy turned 9 recently and we send our love

Off topic: I'm saying goodbye to my almost 14 year old, occasional-toilet-paper-eating dog today. Please send pet pictures. Boxer dogs especially appreciated.