- EvolutionaryScale raises $142 million in a seed round led by Nat Friedman, Daniel Gross, and Lux Capital, with participation from Amazon and Nvidia.
- The funding will support development of ESM3, an advanced AI model for creating novel proteins in drug discovery and materials science.
- ESM3, trained on 2.78 billion proteins, integrates sequence, structure, and function analyses to generate new proteins, similar to Google DeepMind’s AlphaFold.
- EvolutionaryScale plans to offer a 98-billion-parameter model through its Forge platform and a smaller offline version for broader accessibility.
- Applications include creating variants of proteins like green fluorescent protein (GFP) for biological imaging and potential therapeutic discoveries.
- Revenue strategies involve partnerships with pharmaceutical companies and researchers, leveraging ESM3 for commercial breakthroughs.
Main AI News:
EvolutionaryScale has announced a significant funding round, raising $142 million to advance its AI capabilities in generating novel proteins for scientific research. Led by ex-GitHub CEO Nat Friedman, Daniel Gross, and Lux Capital, with support from Amazon and Nvidia’s corporate venture arm NVentures, the startup aims to push boundaries with its latest AI model, ESM3.
ESM3 represents a pioneering leap in biological AI, designed to facilitate protein creation crucial for drug discovery and materials science. Co-founder and chief scientist Alexander Rives highlighted ESM3 as a tool to engineer proteins from first principles, akin to engineering structures and writing computer programs.
Trained on a massive dataset of 2.78 billion proteins, ESM3 integrates sequence, structure, and function analyses to generate new proteins, akin to Google DeepMind’s AlphaFold. EvolutionaryScale plans to democratize access to its 98-billion-parameter model through its Forge developer platform and a scaled-down offline version.
The startup has already demonstrated ESM3’s capabilities by producing a variant of green fluorescent protein (GFP), known for its applications in biological imaging. EvolutionaryScale envisions revenue streams through partnerships with pharmaceutical companies and researchers, leveraging ESM3 to drive breakthroughs in biotechnology.
Looking ahead, EvolutionaryScale aims to integrate ESM3 into AWS’s SageMaker and Nvidia’s NIM microservices, enabling customers to customize and optimize the model with their own data. While profitability may be a long-term goal, the startup remains optimistic about the transformative potential of AI in biology, anticipating broader applications beyond protein design.
With competition from established players like DeepMind’s Isomorphic Labs and others, EvolutionaryScale is poised to scale its AI training efforts, aiming to develop a versatile AI model for diverse biotech applications.
For EvolutionaryScale, backed by prominent investors, the journey ahead promises to redefine the intersections of AI and biotechnology.
Conclusion:
EvolutionaryScale’s substantial $142 million funding infusion, backed by key industry players like Amazon and Nvidia, signifies a pivotal moment for advancing AI-driven protein generation in biotechnology. The introduction of ESM3 not only promises breakthroughs in drug discovery and materials science but also sets the stage for democratizing access to advanced biological modeling tools. With a robust revenue strategy and plans to integrate with major cloud platforms, EvolutionaryScale is poised to lead innovation in the burgeoning intersection of AI and biotech.
