Latent Labs, a new startup founded by a former DeepMind employee, has launched with $50 million in funding. The company aims to make biology "programmable" by building AI-based models and plans to collaborate with biotech and pharmaceutical companies to generate and optimize proteins.
Proteins play a critical role in human biology. They control cellular functions such as enzymes, hormones, and antibodies. Proteins are made up of around 20 different amino acids that link together in sequences, folding into three-dimensional structures, which determine their functionality. In the past, understanding the shape of proteins was a slow and challenging process. DeepMind's AlphaFold model made significant progress in this area, predicting the shapes of over 200 million proteins. This data allows scientists to better understand diseases, design new drugs, and even create synthetic proteins.
Latent Labs is aiming to work in this space by developing AI models that will help researchers create new therapeutic molecules from scratch.
The founder of Latent Labs, Simon Kohl, was previously a key member of the AlphaFold2 team at DeepMind. He later led the protein design team and set up DeepMind's lab at the Francis Crick Institute in London.
Kohl was inspired to start a leaner and more focused startup after witnessing the strategy behind DeepMind's AlphaFold2 and the creation of Isomorphic Labs. In late 2022, he left DeepMind to lay the foundations for Latent Labs and officially registered the company in London in mid-2023.
The company currently employs 15 people, including two former DeepMind employees, a senior engineer from Microsoft, and PhDs from the University of Cambridge. Latent Labs has two main centers: one in London, where AI models are developed, and another in San Francisco, which focuses on testing these models in the real world.
Latent Labs' ultimate goal is to reduce the need for laboratories over time and transform biology into a fully digital field. This could significantly accelerate the drug discovery process and reduce years of experimentation.
The company does not plan to develop its own drugs but aims to collaborate with pharmaceutical and biotech companies to expedite and de-risk their research processes. Latent Labs will either provide its AI models to these companies or support their research through joint projects.
Latent Labs' $50 million funding includes a $10 million seed round and a $40 million Series A round, co-led by Aaron Rosenberg, former strategy head at DeepMind, and Sofinnova Partners.
The funding will mainly be used for employee salaries, acquiring high-performance GPUs for AI model development, and the overall growth of the company. The field where AI and biology intersect is still in its early stages, and it is unclear which approach will prove most effective. However, Kohl believes that Latent Labs has the potential to bring significant innovations to this field.
