Claude designed lab-tested binders against scorpion neurotoxins
Original title: Claude designed protein binders against scorpion neurotoxins
Of 211 designs tested by two CROs, 17 met binding criteria; cellular neutralization was not confirmed.
Anthropic reports that an internal Claude research model, prompted with Citrini Research, designed protein minibinders against scorpion neurotoxins over 64 hours. The model generated 318,484 designs and, with 31 added by the team, 211 were lab-tested by Adaptyv Bio and Twist Bioscience. Seventeen met the authors' binding criteria against four toxins dangerous to humans. Anthropic does not claim an antivenom and released all 211 designs and their data.
Unconfirmed: Binding was lab-confirmed, but neutralization in a cellular environment was not, so antivenom efficacy is unproven.
- Claude produced binders for toxins from campaigns covering four regions.
- The authors say no de novo scorpion-neurotoxin binders had been described before.
- Outside reviewers called the work a promising first step, not a drug.
- Scorpion stings are cited as more than 1.2 million a year, with more than 3,250 deaths.
- Organizations
- AnthropicCitrini ResearchAdaptyv BioTwist Bioscience
- AI models
- Claude
From the original article
10 min read at anthropic.com
Summary: As part of our work experimenting with how Claude can help accelerate drug discovery, we asked an internal research model to design minibinders (small proteins that attach tightly to a target) against scorpion neurotoxins. Claude worked from our prompt and four follow-ups, which we developed in collaboration with the financial investment research firm Citrini Research.
By artninja1988 · anthropic.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.