Company releases open-source code to accelerate collective action in generative biology
Somerville, Mass. – November 15, 2023 – Generate:Biomedicines today announced a publication in Nature describing Chroma, a generative AI model that can program novel proteins not previously found in nature with targeted biophysical, biological, and therapeutic properties. Along with this announcement, the company is making the model available to the broader research community with the goal of accelerating scientific progress in generative biology. The code behind Chroma is available as open-source software and the model weights are freely accessible to academic researchers and non-profit entities.
“What is exciting about this result is that we have demonstrated that fundamentally new ‘molecular concepts’ can now be generated at will and used to create novel proteins with a diverse set of desired properties,” said Gevorg Grigoryan, Co-Founder and Chief Technology Officer of Generate:Biomedicines. “By making the code open source, we’re proud to share this breakthrough in generative biology with the world to help advance our shared goals of fundamentally changing the way protein therapies are created.”
While protein-based therapeutics are a substantial part of the pharmaceutical market, nearly all existing treatment approaches focus on modifying proteins found in nature – which represent only a small fraction of all possible proteins. Through its Chroma model, Generate:Biomedicines can now harness the power of generative AI to infer generalizable principles governing the relationships among protein sequence, structure, and function. In doing so, the company can quickly design novel protein molecules with a wide range of therapeutic properties, realizing the potential for generative AI where protein design developments happen much faster than previously possible – with significant impact on health, materials science, and other areas.
“Nature’s publication of our Chroma model sends an unequivocal message to the broader scientific community that biology is becoming programmable,” said Mike Nally, Chief Executive Officer of Generate:Biomedicines. “By combining machine learning with world-class wet lab capabilities, we are uniquely positioned to continue developing medicines that can treat and cure humanity’s most intractable diseases, all while establishing and sustaining our competitive advantage in AI-enabled drug discovery and development.”
To learn more about the Chroma model, download the open-source package, or read the open access publication in Nature, visit generatebiomedicines.com/chroma.
About Generative Biology
Generative biology represents a fundamental shift in therapeutic development driven by generative artificial intelligence (AI). This approach creates never-before-seen therapeutic molecules targeted to specific biological processes involved in disease that can be modulated with a wide range of protein modalities—from short peptides to complex antibodies, enzymes, and cytokines. But the promise of generative biology goes beyond existing proteins found in nature and can create novel proteins that are purpose-built to address an existing or emerging therapeutic need. As a result, generative biology promises to leave trial-and-error drug discovery methods behind to usher in a new era of programmable drug generation that’s faster, cheaper, and better tailored to specific conditions.
About Generate:Biomedicines
Generate:Biomedicines is the first drug generation company pioneering a machine learning-powered generative biology platform with the ability to create new drugs on demand across a wide range of biologic modalities. The Generate Platform – which is a continuous loop to generate, build, measure, and learn – can drastically increase the speed at which targets and therapeutics are identified and validated. This will improve the specificity of target engagement by generated proteins and reduce the time and cost of identifying and developing clinical candidates. Generate:Biomedicines is a clinical stage company that was founded by Flagship Pioneering after two years of foundational research in its Labs unit and launched in 2020. Learn more about Generate:Biomedicines by visiting https://generatebiomedicines.com/ or following the company on Twitter and LinkedIn.
Generate Media Contact
Megan McLaughlin