Drug discovery often measures biology at the cell level, while therapies must ultimately work across tissues, organs, and whole patients. This scale mismatch means that even highly accurate cellular predictions can fail to translate in the clinic. This session explores strategies to bridge that gap. How do we connect single-cell dynamics to organ-level physiology and patient outcomes? How do we preserve biological context while scaling models? And how do we ensure that virtual biology does not stop at simulation, but informs real therapeutic decisions? Speakers will discuss multiscale modeling that links molecular and cellular systems to higher-order biology; spatial and high-dimensional phenotypic data that retain context; and integrated computational–experimental loops that translate cellular signals into clinically meaningful biomarkers. Together, we ask: how do we ensure virtual biology reflects not just what cells do in isolation, but how biology behaves in the full complexity of patients?

For the first time, AI is enabling us to imagine medicines that “think” - turning on only inside diseased cells or under specific physiological conditions. Neural networks trained on RNA, protein, and cellular data are unlocking a new generation of programmable therapies with unprecedented precision, from cancer drugs that remain inert until encountering tumor signals to RNA medicines capable of adapting to dynamic biological environments. But designing intelligent molecules is only part of the challenge. As AI expands the space of possible therapeutics, the field must also confront a critical question: how do we reliably build, test, and manufacture increasingly complex biological designs? This session explores the emerging continuum from AI-designed molecules to manufacturable programmable therapeutics, examining how advances in sequence design, synthesis, delivery, and validation are translating computational insight into real-world medicines. The future of medicine isn’t static molecules - it’s intelligent, adaptive therapeutics engineered across the full stack, from algorithm to clinic.

The next frontier of biology isn’t in predicting a single static protein structure, but in capturing how proteins move, fold, and interact across time and environments. This session explores how AI can illuminate protein conformations and dynamics, and extend those insights into virtual multi-cellular or tissue models. Experts will discuss the challenge of integrating heterogeneous datasets and instruments, and how breakthroughs in dynamic modeling could reshape drug design, disease understanding, and biomanufacturing. Can we build models that reflect the living, breathing complexity of biology—not just snapshots, but motion?

Why do so many in silico models fail when moved to the lab or clinic? Too often, they’re trained on incomplete, non-human, or non-representative datasets. This session tackles the “data gap” head-on: from interoperability bottlenecks and the black box problem to the limits of current virtual cell simulations (~50 million perturbations vs. the billions biology demands). Panelists will explore how to create “human-first” datasets that reflect real biology, unlock mechanistic interoperability, and close the discovery–development divide. The goal: build AI tools that can directly identify viable drug candidates instead of stalling in silico.

Biology is entering an era where AI agents don’t just analyze data — they co-design, plan, and execute experiments. Multi-agent systems like CRISPR-GPT demonstrate how AI can act as a true lab co-pilot: decomposing complex genome editing projects into stepwise workflows, selecting tools, troubleshooting, and even drafting protocols that allow junior researchers to perform sophisticated edits on their first attempt . Beyond CRISPR, new systems like BioMARS integrate reasoning agents with robotics, while biotech companies are testing “AI lab assistants” that monitor and adjust experiments in real time. This session explores how multi-agent copilots are making biology more reproducible, democratizing complex workflows, and pushing the boundaries of lab autonomy. The central question: when AI can plan, troubleshoot, and validate experiments end-to-end, how should scientists and institutions govern this new power?