PulseAugur
EN
LIVE 15:57:01

Anthropic's Claude AI designs 14/15 disease-targeting proteins autonomously

Anthropic's Claude language models, specifically Mythos Preview and Opus 4.8, have autonomously designed proteins capable of binding to 14 out of 15 targeted diseases. These AI-generated proteins were synthesized and tested in wet labs by Adaptyv Bio and Twist Bioscience, confirming their ability to bind to the disease targets, a critical first step in drug development. While not drugs themselves, the success rate of over 35% on single targets significantly surpasses typical human-led efforts, though figures like Martin Shkreli have downplayed the achievement, highlighting the long road ahead for drug development. AI

IMPACT Demonstrates AI's capability in autonomous scientific discovery, potentially accelerating drug development timelines.

RANK_REASON Research milestone involving AI model autonomously designing proteins for disease targets, with wet lab validation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on dev.to — Anthropic tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Anthropic's Claude AI designs 14/15 disease-targeting proteins autonomously

COVERAGE [1]

  1. dev.to — Anthropic tag TIER_1 English(EN) · Faith & Fact - Marky Mark ·

    Anthropic's Claude Just Designed Real Proteins for 14 of 15 Disease Targets — With No Human Guiding It

    <h2> The Machine Went Into the Wet Lab, and It Did Not Blink </h2> <p>In a laboratory somewhere on this indifferent planet, a language model named Claude was handed fifteen disease targets — fifteen locks — and asked to invent, from nothing, the keys. It succeeded against fourtee…