"The AlphaFold parameters are made available under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0) license" 🙂 github.com/deepmind/alpha… (thanks to @BrianWeitzner for alerting me)

AbLang: An antibody language model for completing antibody sequences biorxiv.org/cgi/content/sh… #biorxiv_bioinfo

biorxiv.org/content/10.110… Introducing AbLang, an antibody-specific language model to generate info-rich antibody representions. AbLang's potential is shown by restoring missing Ab residues with higher accuracy than 'germline filling'. Test it out yourself: github.com/oxpig/AbLang!

Checkout our new review: AI in 3D Compound Design! authors.elsevier.com/a/1eVyo,LqArIa… From DPhil student Tom Hadfield

ABlooper, our open-source EGNN-based approach for antibody Fv modelling, is now published in Bioinformatics. Work led by OPIG DPhil student Brennan Abanades academic.oup.com/bioinformatics…

Some database milestones this week: Thera-SAbDab passed 700 WHO-recognised antibody-based therapeutics (742) while CoV-AbDab passed 5,000 expressed antibodies/nanobodies able to bind to coronaviruses (5033)! Thera-SAbDab: opig.stats.ox.ac.uk/webapps/newsab… CoV-AbDab: opig.stats.ox.ac.uk/webapps/covabd…

Yclon: Ultrafast clustering of B cell clones from high-throughput immunoglobulin repertoire sequencing data biorxiv.org/cgi/content/sh… #biorxiv_bioinfo

Finding important communities in spatiotemporal networks is difficult w/out orthogonal data. DPhil student James Wilsenach uses Hidden Markov Graph Models for unsupervised detection/ranking of communities w/ applications to brain state networks appliednetsci.springeropen.com/articles/10.10… #openaccess

Very happy that our work on generative machine learning for antibody design is now out in mAbs tandfonline.com/doi/full/10.10…. All details are in the preprint thread below. The peer-reviewed version contains more controls and also more comparisons of synthetic and experimental data.

TCR-mimetic antibodies (TCRms) are a transformative new therapeutic modality. We've computationally profiled structures of TCRms/TCRs bound to pMHC to highlight trends in their antigen recognition. TCR-likeness offers a derisking strategy for TCRm design. biorxiv.org/content/10.110…

Our new tool, KA-Search, allows exhaustive yet efficient antibody mining through billions of natural sequences (from OAS and/or your private data), enabling new avenues in function & immunogenicity prediction. Work led by Tobias Hegelund Olsen and Brennan Abanades. biorxiv.org/cgi/content/sh…

Excited to share our most recent piece of work at Oxford Protein Informatics Group (OPIG) : ImmuneBuilder, a set of deep learning based tools to predict the structure of antibodies, nanobodies and TCRs. (biorxiv.org/content/10.110…)

W're hiring! If you are interested in computational protein engineering, reach out to us! #roche #postdoc #antibody #proteinengineering linkedin.com/jobs/view/3488…

Very excited to announce our new preprint, with my PhD work on deep learning to predict change in antibody-antigen binding affinity! Investigating the Volume and Diversity of Data Needed for Generalizable Antibody-Antigen ΔΔG Prediction biorxiv.org/content/10.110… 🧵⬇️

Exciting opportunities in machine learning for protein design. Bonus: you’ll join a vibrant team at the birthplace of therapeutic proteins Roche Deutschland : linkedin.com/jobs/view/3616…

Our most recent piece of work "The Patent and Literature Antibody Database (PLAbDab): an evolving reference set of functionally diverse, literature-annotated antibody sequences and structures" has just been released on bioRxiv (lnkd.in/esgbz-Gg)!!

Wing Ki Wong (Roche Diagnostics GmbH) discusses how integrating advanced language models, bioinformatics tools and antigen-mapped repertoires could accelerate the discovery and optimization of therapeutic antibodies. #AIRRC7 Full agenda: tinyurl.com/airrcmeeting7