Eiji Aramaki

Data-driven drug exploration through deeper real-world text processing

Research Director

Eiji Aramaki

Collaborator

Yoshimasa Kawazoe	Project Associate Professor Graduate School of Medicine The University of Tokyo
Satoko Hori	Professor Faculty of Pharmacy Keio University

Outline

In the age of DX, an enormous amount of medical data is being collected in the form of natural languages, such as technical documents, electronic medical records, and pharmacy reports. We are developing a platform for natural language processing of this vast amount of text data. Our goal is to automatically and massively discover optimal drugs and combinations of drugs that reduce adverse events associated with cancer treatment.

Keiichi Inoue

Creation of automated protein engineering led by AI

Research Director

Keiichi Inoue

Collaborator

Ichiro Takeuchi	Professor Graduate School of Engineering Nagoya University
Kazuhito Tabata	Associate Professor Graduate School of Engineering The University of Tokyo

Outline

In this study, two experimental systems, an automated functional evaluation system and a digital bioassay method, will be constructed to obtain large-scale data on amino acid sequences and functionalities of proteins. In addition, we will link these experimental data to a machine learning system online in real time to realize an automated research loop that automatically plans the next experiment. This will enable the realization of next-generation innovative optogenetics tools.

Wataru Suda

Decoding the host-gut microbiota crosstalk by developing an automated sample collection platform for use in high-resolution time series analysis

Research Director

Wataru Suda

Collaborator

Misako Takayasu	Professor School of Computing Tokyo Institute of Technology
Lena Takayasu	visiting fellow Graduate School of Medicine The University of Tokyo

Outline

Despite the plethora of evidence showing the powerful effects of the gut microbiota in human disease, little progress has been made towards creating revolutionary new therapeutics aimed at preventing or intervening in disease progression. We believe this lack of progress can be attributed to our fragmented understanding of the complex biology underlying host-microbiome crosstalks. In this study, we will aim to elucidate host-microbiome crosstalks by scrutinizing the changes in microbiome compositions over time at high resolution. We will achieve this by automating the process of sample collection. Our goal is to identify universal properties underlying gut microbiome prediction and control.

Hideharu Mikami

Establishment of high-speed, high-dimensional closed-loop optical measurement technology and its applications to neuroscience

Research Director

Hideharu Mikami

Collaborator

Yuichi Iino	Professor Graduate School of Science The University of Tokyo
Masato Taki	Associate Professor Graduate School of Artificial Intelligence and Science Rikkyo University
Teppei Matsui	Professor Graduate School of Brain Science Doshisya University

Outline

We aim to contribute to the next-generation Bio-DX by establishing high-speed, high-dimensional closed-loop optical measurement technology, a novel experimental method that breaks through the limitations of conventional large-scale measurement and analysis. We will develop technologies based on advanced photonics and AI technologies, and apply the developed devices to studies of the nervous system of mice and nematodes to elucidate its mechanisms, which have been difficult to analyze in a realistic amount of time using conventional experimental approaches.

Katsuyuki Yugi

Data-driven Darwinian medicine approach for exploring shared metabolic regulatory networks between seasonal affective disorder and hibernation

Research Director

Katsuyuki Yugi

Collaborator

Yoshihiro Izumi	Associate Professor Medical Institute of Bioregulation Kyushu University
Takahiro Kato	Associate Professor Graduate School of Medical Sciences Kyushu University
Genshiro Sunagawa	Team Leader Center for Biosystems Dynamics Research RIKEN
Tsukasa Fukunaga	Associate Professor Institute for Advanced Study Waseda University

Outline

Seasonal affective disorder and hibernation/torpor-like behaviors have several similar phenotypes. We examine whether or not there are shared metabolic regulatory networks that underlie these two phenomena. We record video clips of mouse behavior and automatically identify their phenotypes using motion analysis AI. Biological samples taken from mice and human patients will be subjected to automated absolute quantitative omics measurements and subsequent trans-omics analysis to reconstruct metabolic regulatory networks of seasonal affective disorder and hibernation/torpor-like behaviors. Divergence dating of molecules that belong to the reconstructed networks will be performed to examine if seasonal affective disorder and hibernation share an evolutionary origin.