Basecamp Research brings EDEN's antibiotic and vaccine design models to Claude Science

Basecamp Research brings EDEN's antibiotic and vaccine design models to Claude Science

Researchers can now design antibiotics and prioritize vaccine targets using a conversational interface after Basecamp Research integrated its EDEN models into Claude, including Claude Science. The integration, announced June 30, 2026, combines the reasoning capabilities of Anthropic's AI workbench for life sciences with biological design tools to move from a target to a shortlist of therapeutic candidates in minutes.

The capability is accessible via Claude Science, Claude.ai, Claude Desktop, Claude Mobile, Claude Code, and Cowork through Anthropic's connectors directory.

This development comes as drug-resistant infections contribute to nearly 5 million deaths per year. While new antibiotics are urgently required, especially for pathogens in lower-income countries where last-resort drugs are difficult to access, the pharmaceutical industry has largely retreated from antibiotic development.

"Microbes have been producing antibiotics and evolving resistance to each other for billions of years,"

Glen Gowers, Co-founder and CEO of Basecamp Research, via PRNewswire

Gowers stated that EDEN learned from this biological history, allowing researchers to design new antibiotics in minutes rather than years.

Experimental Validation and Results

Basecamp Research worked with the Machine Biology Group at the University of Pennsylvania to validate the models. Led by Presidential Associate Professor and Fleming Prize winner César de la Fuente, the collaboration demonstrated that 97% of antibiotic peptides designed by EDEN were active against priority pathogens identified by the World Health Organisation (WHO) during lab testing.

One specific candidate, EDEN-7, was tested in mice infected with multidrug-resistant Acinetobacter baumannii, a pathogen linked to global hospital outbreaks. The candidate showed efficacy similar to a last-line antibiotic. Notably, EDEN-7 was generated zero-shot, meaning the model produced the candidate without iterative engineering or subsequent optimization.

"This collaboration shows how frontier biological foundation models can be paired with rigorous experimental validation to accelerate antibiotic discovery,"

César de la Fuente, Presidential Associate Professor, via PRNewswire

De la Fuente described antimicrobial resistance as one of the greatest existential threats to humanity, noting that partnerships between industry and academia are critical.

Accelerating Vaccine Design

The integration also targets the time-sensitive nature of vaccine development. Traditionally, determining which part of an emerging pathogen to target involves empirical laboratory work that can take months, a delay that often costs lives.

EDEN's vaccine design model identifies proteins most likely to trigger a protective immune response, outperforming other comparable genomic foundation models. Through Claude, researchers can describe a problem in plain language and run a prioritization workflow against a pathogen's genetic sequence, potentially reducing several weeks of research per pathogen into a single conversation.

"The antibiotic crisis and the need for new vaccines are two of the most important public health challenges of our time,"

Jonah Cool, Head of Life Sciences Partnerships and Deployment at Anthropic, via PRNewswire

The BaseData Foundation

Unlike many biological AI models trained on narrow sets of well-studied, cataloged organisms, EDEN is trained on BaseData. This database was constructed through expeditions to more than 200 locations across more than 30 countries, sampling extreme environments such as polar ice, deep-sea sediment, thermal springs, and remote high-altitude plateaus.

These efforts have documented more than a million species new to science, resulting in over 10 billion new genes. Basecamp Research claims this represents roughly ten times the content of every public database combined.

The company utilizes informed-consent and benefit-sharing agreements for all samples. Each sequence is traceable to one of hundreds of country-specific permits, allowing a portion of revenue to return to the communities and countries that steward the biodiversity.

Future Scaling

Basecamp Research intends to scale BaseData 100-fold over the next two years. This goal will be pursued through the Trillion Gene Atlas, a partnership involving Anthropic, NVIDIA, Ultima Genomics, and PacBio. The project is designed to generate genomic data at the trillion-gene scale to further AI-driven drug discovery.