Gingko heads into the Cloud for AI drug development

Large language models capable of generating ChatGPT-style information could accelerate drug discovery and development, says Ginkgo Bioworks as it inks a five-year deal with Google.

The agreement sees Gingko commit to paying Google a minimum of $289 million over five years to purchase cloud hosting services in exchange for various discounts on such services. The firm will use the technology to develop large language models (LLMs) running on Google Cloud’s Vertex AI platform to aid customers accelerate discovery and development in numerous fields, including biopharma.

“The basic idea is that AI tools for biology can enable us to draw new and better insights from large-scale biological data and make predictions for better-performing products and processes for a range of applications,” a Gingko spokesperson told us.

DepositPhotos/MichaelVi

The Vertex AI platform is a managed machine learning platform that provides the infrastructure and tools needed to build and train large language models, while Ginkgo’s Codebase is a massive collection of biological data that can be used to train these models. The partnership aims to combine their strengths to build and train novel foundation and task-specific models for core biological engineering challenges.

Specifically, the spokesperson said, the LLM is trained on a massive dataset of text and code, allowing the model to learn the underlying patterns of language of code, which can then be used to solve problems.

“So for a protein LLM, you can teach it about enzyme function, stability, all sorts of stuff that matter for bio outcomes; for DNA or RNA you can optimize expression, targeting and other factors important for manufacturing and design of programmable therapeutics.”

As such, these models could be used to accelerate the development of new drugs, improve the design of bioengineered organisms, and develop new ways to diagnose and treat diseases.

“Since it can take years and billions of dollars to develop a new drug, and even then, there is no guarantee if it will be successful, this partnership can help accelerate the drug discovery process by automating many of the tasks that are currently done manually,” we were told.

“These models could be used to accelerate the development of new drugs, improve the design of bioengineered organisms, and develop new ways to diagnose and treat diseases.”

Some examples given as to how artificial intelligence (AI) can be used are: