KAIST Develops NUDGE: Single-Stimulus Framework to Redirect Cell Fate

A research team at KAIST has developed a computational framework called NUDGE that can redirect cells toward desired states using only a single, temporary intervention—a departure from existing methods that require permanent genetic changes. The work, published in September in the Proceedings of the National Academy of Sciences, addresses a fundamental problem in regenerative medicine: how to guide a stem cell into becoming a heart cell, or a neuron, or to restore an immune cell to a healthy state, without locking those cells into rigid configurations that leave them unable to adapt.

The challenge with current cell-control technologies is their reliance on permanent intervention. To push a cell toward a target state, researchers typically keep specific genes switched on or off indefinitely. This works, but it carries costs. Cells lose plasticity—the flexibility that allows them to respond to future environmental changes. Worse, permanent control can create abnormal cell states that do not exist in nature. The KAIST team, led by Professor Kwang-Hyun Cho from the Department of Bio and Brain Engineering, took a different approach. Rather than forcing a cell into a fixed configuration, they designed a system that nudges the cell's own internal gene-regulation machinery toward the desired outcome, then lets go. The cell finds its way there on its own.

NUDGE works by mapping the logical relationships between genes inside a cell—which genes activate or inhibit which other genes—and then using mathematics to identify the minimal set of temporary changes needed to shift the cell from its current state to a desired one. The framework can pinpoint exactly which molecular targets need brief activation or inhibition, and it can predict the pathway the cell will follow during the transition and how stable the final state will be. The team tested this logic on 63 published biological network models where a single permanent intervention had been sufficient to reach a target state. In 55 of those cases, the permanent approach had reduced plasticity or created abnormal attractors—states the cell would naturally settle into but that do not occur in healthy biology. NUDGE's approximation method, designed to handle large-scale networks, achieved an average intervention error below 0.01 in more than 90 percent of 552 control problems across 69 biological networks.

To validate the framework against real biology, the researchers applied it to three distinct cellular processes. In cardiomyocyte differentiation—the conversion of a stem cell into a heart muscle cell—NUDGE reproduced a known pattern: the gene MESP1, which is critical for heart development, is naturally active early in differentiation but then turns off in mature heart cells. The model showed how a single temporary intervention involving MESP1 and another regulatory target could guide the entire differentiation process without keeping MESP1 continuously active. In mast cell formation, the team identified GATA2 and GATA1 as key regulators of cell fate. For macrophages, immune cells that can adopt inflammatory or anti-inflammatory states, NUDGE derived optimal strategies for temporary intervention—including one using the signaling molecule IL-4—that could induce anti-inflammatory states while preserving the cell's capacity to switch between phenotypes as needed.

The significance lies in generality. NUDGE is not a tool for one specific cell type or disease. It is a foundational framework applicable across stem cell differentiation, aging, autoimmune disease, regenerative medicine, and cancer reversion therapy—the possibility of returning cancer cells to a near-normal state. The work was conducted by Ferio Brahmana, a master's student, and Corbin Hopper, a PhD student, both in KAIST's Department of Bio and Brain Engineering, with PhD student Woojeong Lee. Professor Cho described the technology as a general-purpose cell-control system that guides cells from their current state to a desired state through a single minimal stimulus. The implications are broad: designing new stem cell therapies, regenerating damaged tissue, modulating immune responses in autoimmune conditions, and reversing the cellular damage of aging. The framework is now available as a tool for researchers designing control strategies across multiple domains of medicine.