Basecamp Research, a London-based AI startup, raised $140 million in funding to advance its biological AI models, called EDEN, which are trained on genetic material from rainforests, oceans, and hot springs. The company plans to use the funds to develop its models and push its therapy candidates toward clinical development.
The funding round, led by investor S32, included participation from Nvidia, Anthropic's Anthology Fund, the NATO Innovation Fund, and Redalpine, among others. Founded in 2020, Basecamp aims to use this capital to expand its research and development efforts in biological AI.
"Given the complexity of biology, we just need orders of magnitude more data," said Philip Lorenz, Basecamp's CTO. He emphasized the importance of combining large-scale data collection with targeted experiments to drive progress in medical applications.
Public genome databases are insufficient for biological AI due to their limited scope. Basecamp's research paper highlights that only 68 percent of the sequence volume in the Sequence Read Archive comes from five species, with humans alone accounting for about 54 percent. This gap underscores the need for more comprehensive data collection.
To address this, Basecamp collaborates with researchers in over 30 countries, collecting samples from under-studied environments such as rainforest soil and the deep sea off Antarctica. The company's dataset now holds about 15 trillion tokens, comparable to the text datasets used to train models like Claude or GPT.
Lorenz sees evolution as a key driver for medical innovation, pointing to bacterial arms races as a source of new drug candidates. By analyzing genetic material from diverse organisms, Basecamp aims to design new antibiotics and tools for cell therapies. The company's latest funding announcement highlights the Trillion Gene Atlas as the foundation for its EDEN models.
Source: thedecoder
