Generate Creates Over 1 Million Proteins With Nvidia/Samsung Investments, Develops Generative Models for De Novo Design
As the "working molecule" of life, the study of the structure and function of proteins requires a lot of time and resources, but the introduction of AI technology has completely changed this situation. Through deep learning models, researchers can quickly predict the three-dimensional structure of proteins, design new functional proteins, and even reveal the hidden connection between genes and diseases. This revolutionary breakthrough has not only attracted close attention from global capital, but also inspired many start-ups to compete and cooperate in the field of AI-driven protein science.
In this competition, Generate: Biomedicines (hereinafter referred to as Generate) stood out and became a leader in the field of AI protein. This technology-driven biopharmaceutical company was founded in 2018.Its unique programmable biology platform,Not only does it deeply integrate artificial intelligence into protein engineering, it also helps scientists design more efficient solutions for traditionally difficult-to-drug targets.
On December 18, 2024, Generate once again attracted industry attention and announced that it had received strategic investment from Samsung Science Life Science Fund. Although the specific amount has not yet been disclosed, the significance behind it is self-evident. It is worth mentioning thatGenerate was also selected for the 2024 CNBC Disruptor 50 list.Become a "dark horse" in the field of AI protein.
Generate：Biomedicines official website：
https://generatebiomedicines.com/
Generate was born from an email
Generate was incubated by the venture capital firm Flagship Pioneering and is headquartered in Massachusetts, USA. In the first two years, Generate focused on basic research in programmable biology, and its unique concept gradually took shape and was officially launched in September 2020.
Company co-founder Dr. Gevorg Grigoryan is a computational biologist specializing in protein structure and function research.Prior to founding Generate, he worked at Dartmouth College, focusing on protein structure and function research.
At the time, he was exploring an innovative idea about designing therapeutic proteins through a data-driven approach, but was confused about how to transform this scientific idea into a practical business. To this end, he dug out the syllabus of the entrepreneurship course he took while pursuing his doctorate at MIT and decided to write an email to Noubar Afeyan, his then-teaching instructor.Afeyan is co-founder and CEO of Flagship Pioneering.He is also a leader in global biotechnology innovation.
Grigoryan spent hours polishing the email and finally sent it in the early morning. Just 15 minutes later, he received a reply from Afeyan. This seemingly accidental contact laid the foundation for the birth of Generate. Under Afeyan's guidance, Grigoryan's idea quickly gained support and attracted more scientists and investors to join. Eventually,Grigoryan, Afeyan and other core team co-founded Generate.
Today, Grigoryan serves as the company's co-founder and CTO, along with Molly Gibson, who is also a founding partner of Flagship Pioneering.
Since its establishment, Generate has been favored by the capital market for its unique technology platform and business model.
In September 2020, the company received a $50 million seed round of financing from Flagship; in November 2021, it completed a $370 million Series B financing; in September 2023, Generate completed another $273 million Series C financing, bringing the cumulative investment to nearly $700 million.Top institutions such as Nvidia, Amgen and Mirae Asset have joined in.It provides a reliable financial foundation for Generate's project development.
Developing Chroma models to design novel protein molecules with therapeutic potential
Unlike traditional pharmaceutical companies, Generate’s core goal is no longer to discover proteins that already exist in nature and have therapeutic potential.Instead, we focus on designing and producing completely new proteins that have never existed in nature.These proteins have potential therapeutic functions.
To achieve this goal, Generate built its proprietary Generative Platform.The platform adopts a "generate, build, measure and learn" cycle model.From target discovery to therapeutic drug design, the speed and efficiency of identification and verification have been greatly improved.
On this platform,Generate has successfully created more than one million proteins that do not exist in nature.Its functions were verified by cryo-electron microscopy and wet experiments. For example, the performance of the green fluorescent protein (GFP) designed by Generate is 50 times better than before.
This research direction had an important breakthrough in November 2023, when Generate published a generative AI model called Chroma in Nature. Chroma provides scientists with a new design tool by inferring the relationship between protein sequence, structure and function.
Paper address:
https://www.nature.com/articles/s41586-023-06728-8
Experiments have shown that the proteins produced by Chroma not only have good biophysical properties, but also have therapeutic potential that can be verified through in vitro and in vivo experiments, and exhibit extremely high stability and functionality in preclinical tests.
More importantly, to accelerate the development of generative biology, Generate open-sourced the Chroma model.
Chroma model open source address:
https://github.com/generatebio/chroma
Generate believes that generative biology marks a fundamental change in drug development driven by generative AI. This approach breaks through the reliance on natural proteins, allowing scientists to design new molecules for current and future treatment needs, accelerating the entry into the era of programmable drugs, making drug development faster, cheaper, and more flexible, thus bringing more hope to patients around the world.
Cooperate with top pharmaceutical companies such as Novartis and Amgen to build multiple drug R&D pipelines
In addition to developing its own products internally, Generate is not limited to "working behind closed doors", but actively cooperates with leading global pharmaceutical companies and biotechnology giants to achieve win-win results.
January 2022Generate has a research collaboration agreement with Amgen toAs part of the research collaboration, Amgen paid $50 million in upfront funding to develop protein therapies for five clinical targets, with the total potential transaction value of the entire collaboration program reaching $1.9 billion. In addition, the company is also working with Roswell Park, MD Anderson Cancer Center, and others to develop protein and CAR-T therapies.
September 2024Generate announced a multi-target collaboration with Novartis.Novartis paid Generate a total of $65 million upfront, including $15 million for Generate's equity. In addition to double-digit tiered royalties, Generate is eligible to receive more than $1 billion in performance-based milestone payments from Novartis.
Currently, Generate has built a rich pipeline covering the fields of tumors, immune-related diseases and infectious diseases. Two of its candidate drugs have entered Phase I clinical trials, including GB-0895, a TSLP monoclonal antibody for asthma, and GB-0669 for Covid-19 PrEP.
Final Thoughts
Generate is redefining the future of protein science and drug development with generative AI. From protein design to clinical trials, the company has demonstrated its unique technical strength and superior execution capabilities.
As AI continues to penetrate the biomedical field, Generate is expected to turn the vision of "designing proteins on demand" into reality, ushering in a more flexible, efficient and precise era of drug development, bringing profound impacts to patients and global health.
References:
3.https://www.cnbc.com/2024/05/14/generate-biomedicines-cnbc-disruptor-50.html
4.https://news.mit.edu/2024/creating-crossroads-gevorg-grigoryan-0628
5.https://www.ourcrowd.com/startup/generate-biomedicines
