A former DeepMind scientist who helped develop AlphaFold, the AI mannequin that revolutionised protein construction prediction, has raised $50 million to launch Latent Labs, a startup centered on utilizing synthetic intelligence to design new proteins.
Simon Kohl, a German-born physicist and machine studying skilled, based Latent Labs in 2023 with the ambition of “achieving computational mastery over biology.” The corporate goals to accomplice with biotech corporations to speed up drug growth by changing conventional experimental strategies with AI-driven protein design.
The funding spherical, accomplished in December, was led by Radical Ventures, an AI-focused funding fund, alongside Sofinnova Companions, recognized for its life-sciences investments. Excessive-profile backers embody: Jeff Dean – Google’s chief scientist, Aidan Gomez – co-founder of generative AI agency Cohere and Mati Staniszewski – co-founder of AI audio startup ElevenLabs.
With an earlier $10 million in start-up capital, Latent Labs has now secured a complete of $50 million to scale its operations.
Kohl, 34, beforehand labored at DeepMind, the place he performed a key position within the second iteration of AlphaFold—expertise that has been recognised with a Nobel Prize in Chemistry. Whereas AlphaFold allows scientists to map present protein constructions, Kohl believes the subsequent step is to design new proteins from scratch utilizing generative AI.
“[AlphaFold] allows you to look at the catalogue of existing natural proteins, but it doesn’t allow you to make new ones,” stated Kohl.
His purpose with Latent Labs is to develop AI fashions that get rid of the necessity for conventional laboratory strategies, making drug discovery quicker and extra exact.
Latent Labs is assembling a staff of prime AI and biotech researchers throughout London and San Francisco, together with 4 former DeepMind scientists. The corporate is actively hiring extra specialists because it expands.
By making use of generative AI to protein and molecular design, Kohl believes Latent Labs can cut back failures in medical trials, a key problem within the pharmaceutical business.
“There are still lots of failures in clinical trials, and in some ways, you can trace the issue back to how the drug was originally discovered,” he defined.
“If you can design proteins and molecules with a higher degree of control and precision, the idea is that in the future, we will see fewer failures.”
Kohl acknowledged considerations over the moral implications of AI-generated proteins, calling it “an important question” that requires broader dialogue. Nonetheless, he pressured that present AI expertise shouldn’t be but superior sufficient to surpass what human consultants can obtain in a well-equipped lab.
As for when Latent Labs’ AI will outperform conventional strategies, Kohl admitted it was tough to foretell.
“Our north star is that our AI systems will eventually be so advanced that they no longer require experimental lab validation,” he stated. “At that point, you will have quite powerful technology on your hands.”
Latent Labs is coming into a aggressive however nonetheless rising subject the place a number of startups are exploring how AI can reshape drug growth. Kohl in contrast the panorama to an area race.
“It feels like quite a number of rockets are being launched right now, and they may all reach orbit,” he stated.
Whereas Latent Labs has not disclosed its present initiatives or shoppers, the startup’s speedy funding and high-profile backing recommend it’s poised to be a key participant in the way forward for AI-driven drug discovery and protein engineering.
