AI Biotech Startup EvolutionaryScale Secures $40M in Funding for Biological Language Models

NNicholas August 26, 2023 12:02 AM

Former researchers from Meta have launched a new AI biotech startup, EvolutionaryScale, securing $40 million in funding. The team aims to build advanced biological language models for applications ranging from cancer cell targeting to environmental clean-up.

EvolutionaryScale: A New Player in AI Biotech

EvolutionaryScale, a biotech startup recently launched by former Meta researchers, is led by Alexander Rives, who was previously in charge of Meta AI’s protein-folding team. This talented team is known for developing a transformers-based AI model that predicts the structures of proteins. The model was trained on protein molecule data and used to create a database of 700 million potential 3D protein structures, which are crucial in drug development, pollution cleanup efforts, and alternative methods of industrial chemical production. With this strong foundation, they now aim to pioneer the development of even more advanced biological language models.

Investors Back EvolutionaryScale’s Vision

In its initial funding round, EvolutionaryScale has secured a significant $40 million from Lux Capital, a venture capital firm known for its investments in emerging tech and science companies. Notably, prominent AI investors Nat Friedman and Daniel Gross also participated in the funding round. This substantial initial investment has set the startup's valuation at a robust $200 million, indicating strong investor confidence in the team's innovative vision for AI-driven biotech.

Pioneering AI Solutions in Biotech

EvolutionaryScale's core objective is to leverage the power of AI in the realm of biotechnology. The team's pioneering vision encompasses a range of applications, from medical to environmental. These include the development of programmable cells that can actively seek out and destroy cancer cells, as well as the design of organisms engineered to clean up toxic waste. By harnessing AI's predictive and analytical capabilities, they aim to push the boundaries of what is currently possible in the field.

While AlphaFold, developed by Google subsidiary DeepMind, currently dominates the protein-folding AI sector, EvolutionaryScale claims its model can make predictions 60 times faster. This could significantly speed up the process of identifying potential protein targets for drug development. However, it's important to note that EvolutionaryScale's model is currently less accurate on average compared to AlphaFold. As the startup continues to refine its AI algorithms, it will need to balance speed with precision to ensure reliable results.

In terms of financial plans, EvolutionaryScale anticipates spending $38 million in its first year, with a significant portion allocated to computing power. The startup's long-term goal is to develop and sell a general-purpose AI model for biology, capable of analyzing a wide range of biotech data. This could revolutionize the industry, as it would allow researchers to apply the model to various applications, whether in medical research or environmental biotech. However, achieving this ambitious goal will require significant resources and technological advancements.

More articles

Also read

Here are some interesting articles on other sites from our network.