Scientists are developing AI models that could simulate how human cells respond to drugs and genetic changes, allowing researchers to test ideas computationally before moving them into expensive laboratory experiments. GenBio AI, a Palo Alto-based startup, is among the companies pursuing this technology with what it calls an AI-driven digital organism that connects biological information across multiple scales.
The company plans to release its first implementation of a virtual cell, according to Axios. Eric Xing, GenBio AI's co-founder and chief scientist, said the technology is envisioned as a simulator in which scientists could introduce a drug or alter a gene and predict how those interventions affect biological systems before running physical experiments.
Beyond single-task models
The approach goes beyond many existing AI models in biology, which solve narrower problems such as predicting protein structures or analyzing individual types of biological data. GenBio's system, called the Artificial Intelligence-Driven Digital Organism (AIDO), is designed to connect information across biological scales - from molecules to proteins to cells - and enable prediction and simulation of cellular processes.
A review published in the British Journal of Pharmacology in July said advances in single-cell data, spatial multi-omics and AI allow researchers to develop virtual-cell systems capable of predicting how biological interventions alter cellular behavior. The researchers identified drug-target prioritization, drug-response prediction and combination therapy design as potential applications.
How virtual cells are built
GenBio is using AI agents to construct the underlying models. In research released this year, its AIDO.Builder system was designed to autonomously select modeling approaches, write and execute code, evaluate results and improve models through repeated iterations. A preprint published on bioRxiv said the system produced competitive results across biomedical benchmarks while reducing manual work normally required to develop predictive models.
The company is not alone in pursuing virtual biology. Researchers associated with the Chan Zuckerberg Initiative previously outlined a roadmap for AI virtual cells capable of representing biological systems across molecules, cells and tissues. A paper available through the National Institutes of Health's PubMed Central described virtual cells as high-fidelity simulations learned from biological data, noting they could help identify drug targets and predict cellular responses.
Limits and obstacles
The technology will not eliminate physical experiments, animal studies or clinical trials. Instead, developers aim to improve early-stage research by narrowing the number of drug candidates and biological hypotheses that need testing in the real world.
Significant technical obstacles remain. A review indexed by the National Library of Medicine identified challenges including integrating different types of biological data, interpreting model outputs and meeting substantial computational requirements. Researchers also need to demonstrate that virtual cell predictions hold up in physical experiments.
Biological responses can vary between tissues and patients, and models trained on incomplete or unrepresentative datasets could produce unreliable results. GenBio is addressing the complexity by combining different forms of biological information rather than relying on a single data type. The company said in June it is working with Nvidia to develop virtual-cell world models capable of simulating human behavior across different biological modalities and scales.
Why this matters for researchers
For scientists working in drug discovery and cellular biology, virtual cells offer a way to prioritize experiments before committing time and funding to wet-lab work. The technology is still facing data integration and reproducibility hurdles, but the early applications - candidate target study, drug-response prediction and therapy design - are directly relevant to research teams seeking to reduce failed experiments. GenBio's use of agent-based model building also points toward lab practices where AI and AI-adjacent AI model development becomes more automated, potentially changing how computational biology teams allocate their time.
Your membership also unlocks:
