Basecamp Research announced the first AI models for programmable gene insertion today, marking a breakthrough in genetic medicine that could enable new treatments for cancer and inherited diseases.
The London and Cambridge, Massachusetts-based frontier AI lab developed the technology in collaboration with NVIDIA, training what it calls the largest evolutionary AI models on a proprietary global dataset. In parallel, NVIDIA's venture capital arm NVentures invested in Basecamp's pre-Series C round following years of technical collaboration.
Programmable gene insertion involves placing large therapeutic DNA sequences at precise locations in the human genome, a goal that has eluded researchers for decades. According to Basecamp Research, current CRISPR-based approaches can only make small edits and require damaging DNA, limiting their therapeutic applications.
Basecamp's AI-Programmable Gene Insertion platform, powered by its EDEN family of evolutionary models, demonstrated insertion at over 10,000 disease-related locations in the human genome. The system produced CAR-T cells showing over 90% tumor-cell clearance in laboratory assays through integration of cancer-fighting DNA into primary human T cells.
The same versatile model designed novel antimicrobial peptides with a 97% lab-confirmed success rate. In collaboration with University of Pennsylvania scientists, the top-performing peptides showed high potency against multidrug-resistant pathogens, offering new tools against antibiotic-resistant superbugs.
The EDEN models trained on over 10 trillion tokens of evolutionary DNA from more than one million newly-discovered species. Data collection spanned five years across 150 locations in 28 countries, creating what Basecamp claims is the largest ethically-sourced biological dataset.
Technical specifications reveal the largest EDEN model trained on 1.95×10²⁴ FLOPS of compute using a cluster of 1,008 NVIDIA Hopper GPUs, accelerated with NVIDIA BioNeMo libraries. This computational scale places it among the most intensive biological models ever reported, comparable to GPT-4 class AI systems. This massive AI compute infrastructure builds on the trend of AI chips powering next-generation computing platforms across industries.
"We believe we are at the start of a major expansion of what's possible for patients with cancer and genetic disease," said John Finn, Chief Scientific Officer at Basecamp Research. "By using AI to design the therapeutic enzyme, we hope to accelerate the development of cures for thousands of untreatable diseases."
The breakthrough addresses a longstanding challenge in genetic medicine with the goal of developing a new generation of curative cell and gene therapies. Basecamp's approach differs from existing methods by using AI to design enzymes capable of performing large gene insertion at defined sites without damaging DNA.
Lab results published today in a paper co-authored by NVIDIA, Microsoft and leading academics show the EDEN models designed multiple active insertion proteins for 100% of tested disease-relevant target sites. The models require only the genomic target site as a prompt, representing a significant advance in AI model capability for biological design.
Basecamp Research collects and curates its own biological data through partnerships with more than 152 organizations in 28 countries. This strategy gives its AI access to genetic diversity not available in public databases, enabling the design of novel protein sequences and biological systems.
The company's work has been recognized with honors including Fast Company's Top 10 Most Innovative Companies in Biotech and the FT-backed Sifted AI100 list of Europe's leading AI startups. Basecamp partners with biopharma companies and academic institutions worldwide.
These capabilities underpin Basecamp's emerging pipeline of cell and gene therapies, opening the path to treatments that are more precise, predictable and personalized than those available today. The company aims to develop potentially curative therapies across a range of cancer and genetic disease indications, representing a new frontier in health technology innovation.
The investment from NVentures will help accelerate Basecamp's research and development efforts as it prepares for its Series C funding round. The deep technical collaboration with NVIDIA spans years and now extends to financial backing through NVIDIA's venture capital division.
The announcement comes as AI applications in biotechnology continue to accelerate, with major tech companies increasingly partnering with life sciences firms. Basecamp's approach of combining proprietary biological data with frontier AI models represents a new direction in therapeutic development.
Industry analysts note that successful programmable gene insertion could transform treatment for thousands of genetic disorders currently considered untreatable. The technology's dual application in both cancer therapy and antimicrobial resistance addresses two of medicine's most pressing challenges.
Basecamp Research's breakthrough demonstrates how AI can move beyond pattern recognition to active design of biological systems. The company's focus on evolutionary patterns in DNA represents a fundamentally different approach from traditional drug discovery methods.
As AI models grow more sophisticated in understanding biological systems, their potential to accelerate therapeutic development increases exponentially. Basecamp's success with both gene insertion and antimicrobial design suggests broad applicability across multiple therapeutic areas, including the kind of health monitoring technologies that could benefit from these medical advances.
The company will continue improving its BaseData dataset, EDEN models and aiPGI platform to advance its pipeline. Basecamp's goal remains developing potentially curative therapies powered by continued advancements in AI-driven biological design.
