Lumen Bioscience’s Post

View organization page for Lumen Bioscience, graphic

5,107 followers

A new paper out with our long-time AI/ML collaborators at @Google and @AAlphaBio: a first-of-its-kind AI/ML method to develop partially variant-proof antibodies for fast-mutating pandemic viruses like SARS-CoV-2. Check it out! https://lnkd.in/gShsqNwk Alone it’s not quite the full solution—in our view, for deep variant protection and high-speed development/deployment, you’ll still need the high-valence, mucosally delivered antibody cocktails only the @LumenBio platform can provide today. But I'm just talking my book there. Sadly, interest in pandemic defense has waned, but there are a few still working hard on it. Hopefully research like this will leave us better prepared the next time. (In our view, AL/ML design has only modest advantages over conventional tech in 'peacetime' drug development.) Tagging a few stalwarts here nonetheless! Bob Nelson at ARCH Venture Partners has always been a big fan of antibodies for pandemic defense. And as luck has it, was talking about AI/ML in clinical development with Jesse Guzman and a crowd at Pioneer Square Labs the night before this preprint dropped. BARDA has been a supporter of our work in this area (though not this particular project), and is one of the few remaining serious investors in pandemic defense, alongside DoD's JPEO-CBRND and Defense Health Agency (AMDRC, MIDRP, etc.). Keep up the good work everyone!

High-throughput ML-guided design of diverse single-domain antibodies against SARS-CoV-2

High-throughput ML-guided design of diverse single-domain antibodies against SARS-CoV-2

biorxiv.org

To view or add a comment, sign in

Explore topics