Yusuke Imoto

Innovation in Interdisciplinary Mathematical Research via Single-Cell Data Science

Grant No.：JPMJCR24Q1

Research Director

Yusuke Imoto

Collaborator

Yuichi Ike	Associate Professor Institute of Mathematics for Industry Kyushu University
Isao Ishikawa	Associate Professor Center for Data Science Ehime University
Kazumitsu Maehara	Assistant Professor Medical Institute of Bioregulation Kyushu University
Kazuyoshi Yata	Professor Institute of Pure and Applied Sciences University of Tsukuba

Outline

This project aims to develop a data analysis framework that identifies the control mechanisms of cell differentiation at the genomic and epigenomic levels via interdisciplinary research between mathematics and biology. This is achieved by integrating advanced mathematical theories that handle high-dimensional spaces, such as high-dimensional statistical analysis, functional space theory, and topology, with the rapidly growing collection of single-cell data worldwide. Furthermore, the project seeks to extend these mathematical techniques to big data analysis across various fields, thereby expanding the possibilities of applied mathematics.

Satoru Iwata

Developing Optimization Methods for Chemical Reaction Prediction, Design and Control

Grant No.：JPMJCR24Q2

Research Director

Satoru Iwata

Collaborator

Taihei Oki	Specially Appointed Associate Professor WPI-ICReDD Hokkaido University
Wataru Matsuoka	Assistant Professor Faculty of Science Hokkaido University

Outline

This project aims at developing a methodology for optimal design of catalyst and reaction conditions by combining quantum chemical calculation and mathematical optimization. To this end, we will introduce a virtual molecular model that approximates and parameterizes electronic and steric effects of molecules. We will then devise an algorithm for optimizing these parameter values with the aid of partial derivatives of the yield or selectivity by each parameter. Finally, we will develop a method to select an actual molecular structure and conditions that approximately realize the obtained optimal parameter values. Thus we will establish a methodology to optimize chemical reactions.

Hiroshi Nishiura

Implementing real time forecasting for disease prevention and control using phenomenological bottom-up models

Grant No.：JPMJCR24Q3

Research Director

Hiroshi Nishiura

Collaborator

Ryosuke Shibasaki	Vice president・Professor Faculty of Engineering Reitaku University
Takemasa Miyoshi	Chief Scientist Cluster for Pioneering Research RIKEN

Outline

We perform forward projection of epidemiological dynamics and causal inference of public health interventions by different effort levels of disease control. Human behavior and climatological variables are also predicted in real time, and those forecasts receive feedbacks from epidemiological models. We aim to achieve real time causal inference of disease control effort in fine spatial and temporal scales, drastically improving life-saving policymaking and promoting comprehensive population-based interventions via the quantitative research foundation of forecasting system.

Atsushi Mochizuki

Controlling and Predicting the Dynamics of Biological Systems Based on Regulatory Networks

Grant No.：JPMJCR24Q4

Research Director

Atsushi Mochizuki

Collaborator

Hideki Ueno	Professor Graduate School of Medicine Kyoto University
Kosuke Yusa	professor Institute for Life and Medical Sciences Kyoto University

Outline

It is thought that biological functions arise from high-dimensional dynamics of network systems involving interactions between many genes. We will elucidate these systems and clarify the principles of biological functions by using (1) RENGE, a method that combines experiments and information science to accurately capture the causal relationships between genes, and (2) Linkage Logic, a mathematical theory that determines key factors based only on causal relationship information. By manipulating the key factors determined by the theory, it is possible to control the dynamics of the entire system. This makes it possible to control cells to any differentiation state. We will also contribute to human health and medicine by controlling the immune system and improving the function of aged immune cells.

Takaharu Yaguchi

Operator Learning Based on Geometric Classical Field Theory and Infinite Dimensional Data Science

Grant No.：JPMJCR24Q5

Research Director

Takaharu Yaguchi

Collaborator

Toshiaki Oomori	Associate Professor Graduate School of Engineering Kobe University
Takashi Matsubara	Professor Graduate School of Information Science and Technology Hokkaido University
Hiroaki Yoshimura	Professor Faculty of Science and Engineering Waseda University

Outline

Recently, the applications of machine learning for accelerating scientific computing have been rapidly growing. In particular, operator learning provides methods for discovering formulas for the solutions of partial differential equations using machine learning, thereby enabling various physical simulations in real time. On the other hand, the theoretical basis of operator learning has not been established. This research aims to develop reliable theories and methods of operator learning by the combination of the theories from geometric classical field theory and infinite dimensional data science.

Aiko Yakeno

Nonlinear, dissipative, and random dynamics based on deterministic governing equations

Grant No.：JPMJCR24Q6

Research Director

Aiko Yakeno

Collaborator

Isao Ishikawa	Associate Professor Center for Data Science Ehime University
Shunsuke Kano	Assistant Professor Mathematical Science Center for Co-creative Society Tohoku University

Outline

Although “flows” follow deterministic governing equations, they are nonlinear, dissipative, and random, making it difficult to understand and control the phenomenon. In recent years, improvements in computational technology have led to further academic discoveries, but there has not been much discussion about the limitations of conventional mathematical science theories when applied to real problems. This research proposal aims to use the latest mathematical science to elucidate and control physical phenomena as real problems, and in the future to apply them to different fields.