- Anthropic is expanding Claude’s role across several stages of drug development
- Generate Biomedicines reports a 70% reduction in CAR T-cell component length
- Modified immune cells produced roughly ten times more cancer-detecting receptors
Anthropic is expanding Claude’s involvement in pharmaceutical research, working with biotechnology developers on projects intended to accelerate the creation of new medicines.
The AI company has laid out in a new blog post how it is examining scientific and administrative obstacles affecting drug development, from molecular design to clinical trial preparation.
Its collaborations involve potential treatments for cancer, food allergies and rare muscular disorders, although most of the treatments remain at experimental stages.
Latest Videos FromTechRadar
Claude enters drug development programmes
At Generate Biomedicines, six scientists used Claude Code to modify a component of CAR T-cell therapy, traditionally employed against blood cancers.
Their work sought to create additional room for genetic instructions that could help the treatment address solid tumours.
The redesigned component was approximately 70% shorter, while laboratory experiments found that immune cells produced roughly ten times more receptors.
The modified cells also destroyed cancer cells more effectively and maintained their activity for longer periods, according to the reported findings.
Divergence Labs used Claude to develop antibody candidates for a potential food-allergy treatment involving the immune molecule IgE.
The system generated approximately 35,000 molecular designs within one month, producing potential molecules for both functions required by the proposed treatment.
Initial experiments identified three designs capable of binding to IgE and another 20 capable of detaching it from immune cells.
Three of the 20 designs appeared more effective than the standard medication at removing IgE from immune cells.
BigHat Biosciences combined Claude with automated laboratory equipment to improve two antibody fragments being considered for future medicines.
Researchers manufactured and tested approximately 1,010 variations over ten weeks, examining binding strength, stability and unwanted molecular aggregation.
The strongest cancer-targeting version bound its intended target more than 10 times as tightly and retained stability at temperatures 18°C higher than the original.
The other fragment, intended for longer-lasting medicines, aggregated more than 100 times more slowly than its original version.
Enveda applied Claude to manufacturing documentation, which had consumed up to 75% of some chemists’ working time.
During an assessment spanning 1,061 pages, the system identified all 11 problems previously discovered by human reviewers, as well as seven additional issues.
The approach reduced manual checking time by 89%, bringing the review process from more than 25 hours to under three.
Anthropic expands into clinical research and drug creation
Anthropic has also partnered with Bristol Myers Squibb to assess clinical study plans and estimate their chances of success.
Michael Pourdehnad, the pharmaceutical company’s senior vice president for early clinical development, said: “From a single prompt, Claude generated a robust initial analysis.”
He explained that this helped evaluate development options that would otherwise have required several days of work.
The companies intend to extend their collaboration further, examining Claude’s potential contribution to decisions during early clinical development.
Anthropic says working directly with drug developers helps identify weaknesses in its models and opportunities for improvement across scientific research.
The company also intends to extend Claude’s involvement through additional partnerships, research programmes and tools for life-science organizations.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
https://cdn.mos.cms.futurecdn.net/n7SCbRu2YRaknUifKEtQEA-1462-80.png
Source link
Efosa Udinmwen




