PARTNERSHIPs

NVIDIA

Advancing generative AI for therapeutic protein design

Evozyne partnered with NVIDIA to develop and refine ProT-VAE, a transformer-VAE architecture that demonstrated unprecedented gains in protein activity and stability. This collaboration brought together deep-learning innovation and high-throughput experimental biology to validate our generative modeling approach.

Impact:

• Established a high-performance modeling backbone used across EvoGen
• Demonstrated significant improvements in catalytic activity and thermal stability
• Provided independent confirmation that generative AI can produce high-function synthetic proteins suitable for therapeutic development

The collaboration underpins our position as an AI-native therapeutics company and supports ongoing advances in engineered proteins for immunology.

Takeda

Applying generative design to rare disease biology

Takeda partnered with Evozyne to explore how generative protein design can unlock treatments in rare, severe genetic disorders. Within six months, the collaboration produced two novel, highly functional proteins and demonstrated EvoGen’s ability to broaden therapeutic options where conventional modalities fall short.

Impact:

• Delivered design-ready proteins on an accelerated timeline
• Demonstrated reproducible gains across multiple protein families
• Reinforced the advantages of engineered proteins in diseases underserved by small molecules and antibodies

This work showed how Evozyne’s platform can rapidly generate new therapeutic starting points in areas of high unmet need.

Bill & Melinda Gates Foundation:

Engineering proteins beyond evolutionary limits

Supported by the Gates Foundation, Evozyne engineered synthetic proteins with significantly improved thermotolerance, as published in microPublication Biology. The project served as an early validation of our iterative design-build-test cycle and demonstrated how generative AI can exceed the constraints of natural evolution.

Impact:

• Created synthetic proteins with superior stability and function
• Generated peer-reviewed evidence supporting platform accuracy
• Produced foundational datasets that strengthened EvoGen’s predictive performance

This work helped establish the scientific basis for applying generative AI to complex immune-driven diseases.

Why these partnerships matter

Each collaboration demonstrates that Evozyne’s engine performs under real-world conditions and across diverse biological challenges. These efforts strengthen the platform, accelerate development in immune-mediated disease, and expand the therapeutic design space.

Evozyne continues to explore strategic collaborations with organizations seeking novel mechanisms in immunology, protein engineering, and next-generation biologics.