The University of Tokyo establishes new theory for controlling uncertain biological populations from molecular reactions to infectious diseases

How can we suppress the spread of infectious diseases? Can we inhibit the proliferation of cancer cells? How can we maintain the balance of diverse species in ecosystems and gut environments? All of these questions connect to the problem of how to control "the movement of uncertain populations." A research group consisting of Doctoral Student Shuhei A. Horiguchi of the Graduate School of Information Science and Technology at the University of Tokyo (at the time of the research, currently Assistant Professor at the Nano Life Science Institute, Kanazawa University), and Professor Tetsuya J. Kobayashi of the Institute of Industrial Science at the University of Tokyo has constructed a theoretical framework for optimally controlling the behavior of biological populations that fluctuate uncertainly. Their work was published in PRX Life.

The research team developed a new theory for more accurate control of the fluctuations of complex and uncertain biological populations. In this theory, by measuring the cost of control using the f-divergence and its special form, the Kullback-Leibler divergence, both of which are introduced in information theory, the team can derive optimal strategies in a computationally easier manner while maintaining generality and adhering to reality. Notably, control strategies that maximize the inherent fluctuations of populations can be obtained, making this a highly versatile method applicable to control problems of diverse biological populations with large fluctuations ranging from molecular reactions to infectious diseases, spanning from microscopic to macroscopic scales.

Specifically, the theory was applied to models for efficient methods to prevent species extinction in environments where multiple species compete, and strategies to effectively suppress the spread of infectious diseases. The results revealed that for models exhibiting exponential growth and decline, not only strategies that "continuously control" but also strategies that "switch control on and off according to the difficulty and effectiveness of control" naturally emerge.

For example, when controlling competing species, it was found efficient to deliberately do nothing when both species exist in similar numbers, and to intensively control when the population of either species decreases. The fact that such "mode switching" automatically emerges from the theory can provide useful guidance for actual social implementation.

By integrating the different fields of control engineering, information theory, and biology, the study provides a powerful theoretical foundation for approaching real-world complex problems flexibly and efficiently. Future applications are expected in various situations, including optimization of cancer immunotherapy, formulation of countermeasures during pandemics, and design of artificial cells and biorobots.

Horiguchi commented: "I was fascinated by how cells autonomously create order as a population and began this research with a desire to understand the mechanism. As our research progressed, I realized that the perspective of mathematically capturing population behavior is useful not only for cells but also for understanding and controlling different phenomena such as chemical reactions, ecosystems, and infectious diseases. In the future, I want to go beyond external manipulation to elucidate how systems achieve intrinsic control and develop that knowledge as a theoretical foundation supporting future applications."

Journal Information
Publication: PRX Life
Title: Optimal Control of Stochastic Reaction Networks with Entropic Control Cost and Emergence of Mode-Switching Strategies
DOI: 10.1103/zttn-tpzq