This advanced press report was prepared to cover the behind-the-scenes details of the invention of the first AI model of the human cell and related topics in the global technology and regulatory sector.
Article index:
- Launch of the first global model simulating the human cell
- Engineering architecture and drug response simulation
- Accelerating research, drug discovery, and disease treatment
- Frequently asked questions
Launch of the first global model simulating the human cell
Palo Alto, California-based biotech startup GenBio AI has announced the launch of its groundbreaking “IdoCell” system, which it describes as the first integrated virtual “foundation model” capable of simulating the precise biological behavior of a human cell across its entire hierarchy and biological lifecycle—starting from DNA and RNA, through complex proteins, and reaching the comprehensive dynamics of the entire cell, both in its natural state and when responding to therapeutic drugs and external stimuli.
The promising company was founded by biochemist David Baker, recipient of the 2024 Nobel Prize in Chemistry, in collaboration with prominent artificial intelligence scientist Eric Xing. The company positions “IdoCell” as a qualitative leap far beyond previous tools such as Google DeepMind’s “AlphaFold,” which is limited to predicting the structures of individual, isolated proteins; the new system aims to model the molecular machinery of the entire cell as an integrated, interconnected system.
Baker stated in remarks to STAT News: “With AlphaFold, you were predicting the structure of a single house or identifying a single individual in an entire city, whereas with the IdoCell model, you are predicting the comprehensive way that city functions with all its details and daily activity.”
Engineering architecture and drug response simulation
The “IdoCell” system relies on a foundation model architecture that makes it multi-scale biologically and capable of preserving the cumulative state of previous cellular interactions. This allows researchers to perform modifications and interventions at any level within the cell—such as disrupting a specific gene or modifying an enzymatic pathway—and observe how cumulative effects propagate and flow from genetic regulation to protein behavior and ultimately to macroscopic changes in cell function.
Founders Eric Xing, Eran Segal, and Lee Song published a research paper in the prestigious journal Nature Medicine detailing their three-phase methodology: building specialized foundation models for each biological modality, interconnecting these models across various microscopic levels, and then aligning the entire cellular network into a cohesive, unified system.
In an initial practical demonstration, the company used the model to simulate the effect of the drug imatinib on leukemia cells, and the system successfully reproduced the precise and well-known mechanism of the drug’s action across multiple biological levels. The system currently supports two continuous human cell lines: myeloid leukemia cells and liver tumor cells, with plans to add new cell lines in the coming months.
Accelerating research, drug discovery, and disease treatment
The company presents its model as an ultra-fast virtual testing and experimentation platform for the pharmaceutical research sector; statistics indicate that only one chemical compound out of every 10,000 drug compounds entering development successfully reaches actual clinical use in patients.
Ziv Bar-Joseph, co-founder and chief scientific officer of the company and former head of AI research at global pharmaceutical firm Sanofi, stated: “A system like IdoCell allows medical companies to evaluate hundreds of drug candidate compounds with extreme precision and speed, paving the way for the development of safer and more effective treatments years ahead of what was previously possible.” The company is preparing to launch an academic collaboration program giving scientists at universities and biotech companies early access to the system.
Frequently asked questions
Question: Who is the Nobel laureate who co-founded GenBio?
Answer: He is biochemist David Baker, recipient of the 2024 Nobel Prize in Chemistry, in collaboration with artificial intelligence scientist Eric Xing.
Question: What is the fundamental difference between the IdoCell model and the AlphaFold model?
Answer: AlphaFold predicts the structures of individual proteins, whereas IdoCell simulates the operation of the human cell with all its molecular machinery as an integrated biological ecosystem.
Question: What cell lines does the model support in its initial version?
Answer: It supports myeloid leukemia cells and liver tumor cells, with work underway to include additional cell lines soon.
Question: How does the virtual model help reduce the costs of drug research?
Answer: It allows for the virtual testing of thousands of drug compounds and their genetic effects, reducing failure rates, laboratory experimentation costs, and years of development.
