RC RANDOM CHAOS

Anthropic Scales AI Drug Development Work With Six Partners

· via Hacker News

Original source

Early progress on accelerating drug development with Claude

Hacker News →

Anthropic’s life sciences team has initiated partnerships with six biotechnology organizations—BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines, and The Column Group—while launching its own drug discovery programs. Rather than focusing AI narrowly on molecular design, the team’s strategy targets bottlenecks distributed throughout drug development: target identification, molecule optimization, manufacturing compliance, regulatory filings, and clinical trial administration.

Early results demonstrate Claude’s utility across these varied tasks. Generate Biomedicines used Claude to redesign genetic instructions for CAR T-cell cancer therapy, producing a variant 70% shorter that proved 10 times more effective in lab tests. Divergence Labs leveraged Claude to autonomously generate roughly 35,000 antibody designs over a month for a novel food allergy treatment, identifying promising candidates for both binding and immune-cell detachment. BigHat Biosciences iterated on drug leads through roughly 1,010 variants in 10 weeks, achieving tenfold improvements in binding and dramatic gains in stability. Enveda deployed Claude to review manufacturing documentation, catching all 11 issues found by human reviewers plus seven additional problems, while reducing review time by 89%. The Column Group is establishing Pleiades, a new company designed to demonstrate how Claude can reduce scientific and administrative overhead at scale.

Anthropic views these partnerships as threefold opportunities: to identify and address limitations in Claude’s current capabilities, to generate case studies for the broader biotech industry, and to contribute directly to new treatments for diseases including muscular dystrophy, food allergies, and cancer. The work signals a broader shift toward AI involvement across the entire drug development pipeline rather than isolated applications.

Read the full article

Continue reading at Hacker News →

This is an AI-generated summary. Read the original for the full story.