GenBio AI Unveils AIDO Cell, First World Model Simulating a Human Cell Across the Full Biological Hierarchy
核心洞察
GenBio AI (搜索) introduced AIDO Cell (搜索), described as the first virtual cell world model simulating human cell behavior from DNA and RNA through protein to whole-cell level.
The system is stateful and multi-scale, allowing researchers to run sequences of perturbations and observe cumulative effects as in a wet-lab workflow.
In an early case study, AIDO Cell (搜索) recapitulated the mechanism of imatinib in leukemia (搜索) cells, tracing drug effects through the cellular environment.
GenBio AI (搜索), an AI for Science company co-founded by 2024 Nobel Laureate in Chemistry David Baker and AI scientist Eric Xing, has introduced AIDO Cell (搜索), a virtual cell world model it describes as the first system capable of simulating a human cell — both in its natural state and in response to drugs and other interventions — across its full biological hierarchy, from DNA and RNA through protein to the whole-cell level.
The Palo Alto-based company said full-hierarchy virtual cell simulation has been widely identified as a key milestone in computational biology. Compared with previous efforts by various research labs, AIDO Cell (搜索) is the first to integrate the entire molecular-to-cellular spectrum into one continuous simulation, according to GenBio. Using the system, a researcher can perturb the cell at any level and observe how the entire cell responds — for example, knocking out a gene and watching effects cascade from gene regulation to changes in protein behavior to effects on the cell as a whole.
"What's exciting here isn't that we've solved cellular biology — we haven't, at least not yet. It's that, for the first time, we have a system capable of simulating a cell across the full hierarchy of biological scales, from DNA to whole-cell behavior, in one place, which lets you interrogate it computationally," said Baker, who received the 2024 Nobel Prize in Chemistry for his work in computational protein design. "AIDO Cell (搜索) is a meaningful foundation to build on, even at this early stage, and could have major implications for fields like drug discovery, toxicology, and disease research broadly."
Stateful Architecture Enables Sequential Perturbation
GenBio attributes AIDO Cell (搜索)'s multi-scale capability to an underlying world model-based architecture, which it describes as one of the first biological applications for an emerging area of AI research.
"What world model architecture enables, and what makes our system genuinely new, is that it's both multi-scale and stateful. The simulation remembers what you've done to the cell. You can run entire sequences of perturbations, where each one builds on the last, and see the evolving, cumulative effect in the cell, just as you would in a real (wet lab) workflow," said GenBio AI (搜索) President and Chief Scientist Eric Xing.
Early Case Study Reproduces Imatinib Mechanism
GenBio framed the system as a response to the cost and attrition of drug development. According to the company, for every 10,000 compounds that enter the drug development pipeline, only one on average reaches the clinic, a process taking over 10 years at an estimated average cost of more than US$2 billion per successful drug candidate. A major reason, GenBio said, is the complexity of human biology and the inability of existing research tools to capture more than a fraction of it.
In an early case study, the system successfully recapitulated the mechanism behind imatinib in leukemia (搜索) cells, demonstrating its ability to trace how a drug treatment cascades through the full cellular environment.
"We've released this version of the system as a preview, an early but functional demonstration that multi-scale, stateful virtual cell simulation is now possible," Xing said. "We plan on releasing more advanced versions of the system with enhanced capabilities and accuracy later this year and over the next year."
Cell Line Coverage and Access Plans
AIDO Cell (搜索) supports K562 and HepG2 cell lines, with additional cell types in development. GenBio said it is preparing to launch an early access and academic collaborator program for scientists across academia, biotech, and pharma.
The company was founded in 2024 and is headquartered in Palo Alto, California, with labs in Paris and Abu Dhabi. Its stated mission is to make biology fully computable through an AI-Driven Digital Organism (AIDO) intended to simulate living systems and accelerate medical and scientific research; AIDO Cell (搜索) is described as the first step toward that larger mission.
GenBio's founding team includes Baker; Ziv Bar-Joseph, co-founder and chief scientific officer and a professor at Carnegie Mellon University and former head of AI at Sanofi; Fred Hu, co-founder and CEO; Emma Lundberg, co-founder and chief strategic advisor and co-director of the Human Protein Atlas; Le Song, co-founder and chief technology officer; and Xing. Eran Segal and Debora Marks serve as scientific fellows.