The race to build a virtual cell
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Sarah Grillo/Axios
A new "digital organism" could simulate how a potential drug or genetic change moves through molecules and cells and eventually people, according to a new Nature Medicine paper out Thursday.
Why it matters: Better AI simulations could streamline biology and help researchers screen ideas earlier and focus experiments on the most promising candidates to catch risks sooner.
The big picture: The scientists describe the project as a kind of flight simulator for medicine. Scientists will be able to run experiments on biology, like altering a gene or introducing a new drug to a cell, before attempting those experiments in the real world.
How it works: Existing biology AI generally handles one specialized task, such as predicting a protein's shape. In the Nature article, Eric Xing and his co-authors argue that the next step is connecting those systems — models for DNA, proteins, cells, tissues and health outcomes — into a dynamic simulator.
- "It is really about building a simulator of biology," Xing told Axios.
- Xing said his company, GenBio AI, will soon release "the world's first implementation of the virtual cell."
Zoom in: In addition to his role as chief scientist and co-founder of GenBio, Xing is president of the Mohamed bin Zayed University of AI in Abu Dhabi and a professor of computer science at Carnegie Mellon University in Pittsburgh.
Zoom out: GenBio is one of several efforts chasing virtual cells.
- BioHub, the institute backed by Mark Zuckerberg and Priscilla Chan, is building an open platform of AI models to understand and predict cell behavior. Its virtual-cell program aims to model cells across tissues, people and conditions with the audacious goal of curing all disease.
- Xing calls Zuckerberg's Biohub a "friendly competitor," but says their approach differs.
- He characterizes Biohub's work as leaning more heavily on language-model-style sequence models, while GenBio is trying to combine sequence, structure, imaging and interaction data in one multi-scale system.
Reality check: Researchers aren't close to building a virtual human.
- The near-term vision is a "virtual cell," a model that can take multiple kinds of biological data into account and predict the effects of a series of interventions, rather than a single static outcome.
- The goal is not to replace animal studies or clinical trials. It's to use AI as a filter: eliminate weaker drug ideas before researchers spend years and large sums testing them.
Yes, but: Biased or incomplete training data, different responses across patients, and AI's tendency to confuse correlation with causation all present challenges to the process.
What we're watching: How quickly virtual cells get built and whether the systems can make predictions about a drug target or side effect that are later confirmed in real-world experiments.
