Latent Labs, a startup founded by former Google DeepMind scientist Simon Kohl, has emerged from stealth mode with a substantial $50 million in funding to pursue its ambitious goal of making biology programmable. The company aims to develop AI foundation models that can generate and optimize proteins, partnering with biotech and pharmaceutical companies to accelerate drug discovery.

The concept of making biology programmable may seem daunting, but it's essential to understand the crucial role proteins play in human biology. Proteins are the building blocks of living cells, driving everything from enzymes and hormones to antibodies. However, determining the shape of each protein was historically a slow and labor-intensive process. This was the breakthrough achieved by DeepMind's AlphaFold, which combined machine learning with real biological data to predict the shape of around 200 million protein structures.

Armed with such data, scientists can better comprehend diseases, design new drugs, and even create synthetic proteins for novel use cases. This is where Latent Labs enters the picture, with its mission to enable researchers to computationally create new therapeutic molecules from scratch. Kohl, who worked with the core AlphaFold2 team and co-led the protein design team at DeepMind, saw an opportunity to create a leaner, more focused outfit specifically dedicated to building frontier models for protein design.

Latent Labs has already assembled a team of 15 employees, including two from DeepMind, a senior engineer from Microsoft, and PhDs from the University of Cambridge. The company operates out of two sites: one in London, where the frontier model development takes place, and another in San Francisco, which houses a wet lab and computational protein design team. This setup enables Latent Labs to test its models in the real world and gather feedback to refine its approach.

Kohl's vision is to ultimately negate the need for wet labs, making biology programmable and reducing the reliance on biological experiments. This would revolutionize the drug-discovery process, which currently involves countless experiments and iterations that can take years. Imagine a world where researchers can design custom molecules without relying on wet labs, and Latent Labs' models can generate protein drugs with desired properties in a "push-button" way.

In terms of business model, Latent Labs doesn't plan to develop its own therapeutic candidates in-house. Instead, it will work with third-party partners to expedite and de-risk the earlier R&D stages, providing direct access to its models or supporting discovery programs via project-based partnerships. The company's $50 million funding includes a previously unannounced $10 million seed tranche and a fresh $40 million Series A round co-led by Radical Ventures and Sofinnova Partners.

The funding will be used to cover salaries, including new machine learning hires, as well as infrastructure costs, such as compute resources required for building large models. With this investment, Latent Labs is well-positioned to double down on its mission, acquire compute resources, scale its teams, and establish partnerships with biotech and pharma companies.

While Latent Labs is not the only venture-backed startup exploring the intersection of computation and biology, Kohl believes that the field is still in its early stages, and there is ample opportunity for innovation. With its focused approach and significant funding, Latent Labs is poised to make a significant impact in the years to come.