AI is increasingly getting used in protein design, the event of latest proteins with specific goal properties. The applications of protein design are diverse, but it surely is a promising avenue for locating drug treatments to fight diseases and developing latest products for home care, agriculture, food and materials.
One of the various providers developing AI technology for protein engineering, Arzedawas founded in 2009 by a gaggle of researchers on the University of Washington's Baker Lab. After constructing the corporate on their very own for nine years, Arzeda's founders – Alexandre Zanghellini, Daniela Grabs, Eric Althoff and David Baker – decided to bring the Seattle-based company's protein design platform to market and attract outside enterprise capitalists.
“It's difficult to go from lab scale to business scale in industrial biotechnology and plenty of firms have did not accomplish that,” Arzeda CEO Zanghellini told TechCrunch. “It's a testament to our progress that we were in a position to raise capital in a difficult environment where many firms within the space needed to exit of business.”
Zanghellini says Arzeda's technology combines “biophysics-based” AI models with generative AI methods, including large language and diffusion models. To train these models, Arzeda uses a proprietary dataset of protein sequences and structures that the startup assembled itself.
Arzeda is just not focused on biomedical applications, but on redesigning chemical-heavy products to make use of alternative, supposedly more sustainable, AI-engineered proteins and enzymes. The startup's first product is a stevia-based natural sweetener for an unnamed consumer brand; other products within the pipeline include laundry detergent for Unilever and biodegradable materials co-developed with WL Gore.
Arzeda also handles the validation and manufacturing of the proteins and enzymes it develops and makes money each from the sale of its proteins and enzymes and from the sale of the ultimate products it produces with contract partners.
“We have developed filters based on our understanding of structural biology and biophysics to further select designs before they’re experimentally tested,” said Zanghellini. “All of our designs are experimentally tested, which allows us to weed out hallucinations.”
Arzeda's strategy has proven successful. Zanghellini told me that the 70-person company is generating revenue and that, along with Unilever and Gore, Arzeda's clients include AAK (a candy fat maker) and the Department of Defense. He declined to say more concerning the Department of Defense deal, apart from that it “has shown very promising results” and “has generated lots of interest.”
The growth helped Arzeda close an oversubscribed $38 million funding round led by Sofinnova Partners and with participation from Fall Line Capital, Sucden Ventures, Silver Blue, Gore's corporate enterprise division, Continental Grain Company, Bunge Ventures and Lewis & Clark AgriFood. Zanghellini said the proceeds, which brought Arzeda's total raised to $83 million, will go toward increasing production of Arzeda's natural sweetener and commercializing other enzymes in the corporate's portfolio.
“Our syndicate brings together reputable investors from the economic and agricultural biotechnology sectors with the essential expertise, network and capital to support Arzeda’s growth,” said Zanghellini. “The combination of this latest capital round with our revenue growth provides a snug runway and a transparent path to ‘corporate-EBITDA-positive.'”
