Masahito Ikawa

Automatic Optimization of Testicular Tissue Culture Using Machine Learning and Its Application to Spermatogenesis Research

Grant No.：JPMJCR21N1

Research Director

Masahito Ikawa

Collaborator

Takehiko Ogawa	Professor School of Medicine Yokohama City University
Hiroshi Kimura	Professor Micro/Nano Technology Center Tokai University
Akira Funahashi	Professor Faculty of Science and Technology Keio University

Outline

We will Construct an automatic quantitative evaluation system for in vitro spermatogenesis using in vivo spermatogenesis as an absolute control. We will improve the in vitro spermatogenesis system by collecting and evaluating a large amount of image data through automated observation of microtubules in culture and repeated Bayesian optimization of experimental conditions. In addition, we will combine interventions such as genetic modification and drug administration to elucidate the essence of spermatogenesis.

Jun Ishii

Data-driven next-generation evolutionary breeding of microorganisms

Grant No.：JPMJCR21N2

Research Director

Jun Ishii

Collaborator

Fumio Matsuda

Professor Graduate School of Information Science and Technology The University of Osaka

Outline

In this study, we will create a data-driven synthetic evolution of microorganisms that makes use of in silico design / in vitro selection in parallel. To this end, the accuracy of computer design for metabolic models should be improved by acquiring experimental data as comprehensively as possible using our robotics-based technologies for microbial construction / selection and evaluation. Ultimately, we aim to establish a selection method for highly fine-tuned, synthetic evolved cells and to understand the essential principles of complex metabolic controls.

Mariko Okada

Development of cell fate control method by natural language processing and computational simulation

Grant No.：JPMJCR21N3

Research Director

Mariko Okada

Collaborator

Hidetoshi Shimodaira	Professor Graduate School of Informatics Kyoto University
Makoto Taiji	Deputy Director・Team Leader Center for Biosystems Dynamics Research RIKEN

Outline

We will develop a computational tool for building mathematical models using natural language processing, integrating molecular and cell simulation, and drug design using machine learning for controlling cancer networks. Through the generation and rejection of hypotheses in cell regulation, we will build a methodology that enables cell proliferation control in the shortest path or in unexpected ways. This will contribute widely to the digital transformation of biological research.

Daisuke Kiga

BioDOS: Uncover what bio-reaction network could be

Grant No.：JPMJCR21N4

Research Director

Daisuke Kiga

Collaborator

Kazuteru Miyazaki	Professor Research Department National Institution for Academic Degrees and Quality Enhancement of Higher Education
Masayuki Yamamura	Professor School of Computing Institute of Science Tokyo

Outline

Beyond human cognitive biases, we will design “what life could be” by constructing Bio Discovery OS (BioDOS), which combines deep learning AI and logical reasoning AI. In addition, we will show that the designed gene network can operate in a variety of organisms and culture conditions. The results will lead to the discovery of new forms of life through another round of search in nature.

Shinsuke Sando

Data-driven science for the creation of cell membrane permeability of medium-sized molecule

Grant No.：JPMJCR21N5

Research Director

Shinsuke Sando

Collaborator

Koji Umezawa	Assistant Professor Graduate School of Science and Technology Shinshu University
Issei Sato	Professor Graduate School of Information Science and Technology The University of Tokyo

Outline

Medium-sized biomolecules, such as peptides, are functional molecules that are expected to be applied for next-generation pharmaceuticals. In this project, we will conduct data-driven research to predict and elucidate passive cell membrane permeability of middle-sized biomolecules, by integrating large-scale experimental data on membrane permiability, molecular simulations, and machine learning. Ultimately, we aim to construct an algorithm for designing membrane-permeable middle-sized biomolecules.

Itoshi Nikaido

Elucidating the Mechanisms of Genome Resilience

Grant No.：JPMJCR21N6

Research Director

Itoshi Nikaido

Collaborator

Martin Frith	Professor Graduate School of Frontier Sciences The University of Tokyo
Totai Mitsuyama	Research Team Leader Artificial Intelligence Research Center National Institute of Advanced Industrial Science and Technology (AIST)

Outline

We will elucidate the mechanisms of genome resilience, which is the ability to protect itself from various perturbations that mutate DNA, using single-cell multi-omics, robotics, and artificial intelligence technologies. From massive data on genomic resilience, we will understand the universal mechanisms by which genomic instability occurs in specific genomic locations in different tissues and cell types.