Skip Navigation

Results and Future Plans

Computational and Experimental Systems Biology Group
Towards an understanding of systems of life

Makoto Taiji

Group Director, Computational Systems Biology Research Group, RIKEN Advanced Science Institute

In addition to working as Deputy Project Director of the Computational and Experimental Systems Biology Group, Makoto Taiji doubled as leader of the High-Performance Molecular Simulation Team. He has developed computers specialized for scientific simulations since his student days.

The three teams of the Computational and Experimental Systems Biology Group, namely, the High-Performance Molecular Simulation Team, Cellular Systems Biology Team, and Omics Bioinformatics Team, are aiming to gain an integrated understanding of all various levels of life systems, using information science, computing science and various experimental approaches.
The group is putting particular focus on life science research using high-performance computers to carry out large-scale simulations and massive information processing, in addition to their investigation into information technologies that will correspond to the high throughput experiments seen in recent years. They are also carrying out research into systems biology that combines computational life science and experimental biology with the goal of practical application.

High Performance Molecular Simulation: The World's Fastest

The High-Performance Molecular Simulation Team aims to understand the behavior of proteins at the atomic level using molecular simulations, and to use molecular designs obtained through simulations to contribute to life science. In order provide the heavyweight computing power necessary for carrying out practical simulations, the team began development of a computer known as MDGRAPE-3. This computer is specialized for molecular dynamics calculations, and its development was completed in 2006 (see photograph). Using this computing system, a theoretical performance of one petaflops (one thousand trillion calculations per second) was achieved for the first time anywhere in the world.
Practical application to the simulation of the aggregation process of the yeast protein Sup35 won MDGRAPE-3 the Gordon Bell Prize in 2006 (Peak Performance Category, Honorable Mention). The team now collaborates with internal and external experimental researchers to carry out simulations using this system for a wide range of applications, including molecular design for designing drugs, analysis of protein function, and analysis of the protein folding process. One of the team's collaborative efforts is their work with the Cellular Systems Biology Team. This work involves calculation of interactions between proteins involved in signal transduction systems, in order to gain fundamental knowledge about systems biology at the molecular level.

MDGRAPE-3 System

The magnificent sight of 138 servers lined up on the 5th floor of the West Research Building of the Yokohama Research Institute.

Algorithms for Analysis of Gene Function and Networks

Elucidation of Systems of Cellular Control

The team is currently using experimental data and mathematical models to analyze intracellular signaling networks, focusing on receptor tyrosine kinase (RTK) and other molecules involved in carcinogenesis. Cells carry out dynamic fate determination like growth or differentiation in response to external stimuli, while maintaining stability. Mathematical analysis helps us to clarify hidden mechanisms that control cell fate.

To date, the Cellular Systems Biology Team, targeting the ErbB receptor signal transduction pathways, has carried out computational and experimental analysis on network kinetics, gene expression control, network structures, and molecular interactions. They have also developed methodologies based on information science. By high integration of computational and experimental methods, the team managed to show that early stage gene expression brought about by the activation of the ErbB receptor is controlled quantitatively, not qualitatively as was previously thought, and that this quantitative control is determined by the ErbB receptor.
It was also shown that crosstalk and feedback in early stage gene expression and signal transduction can transform mere quantitative differences in gene expression into biphasic qualitative differences, which can determine cell fate. This kinetic model of ErbB4 receptor signaling is now widely used as a benchmark by overseas researchers.
Also the team has developed software to predict structures and parameters of unknown signal pathways using genetic algorithms for network analysis. In 2005, the group set up an international Receptor Tyrosine Kinase (RTK) consortium composed of theoretical and experimental signal transduction researchers. The group contributes to international workshops of the consortium which are held once or twice a year. The group maintains a high research standard through collaboration and information exchange.

Creating New Systems through Bioinformatics

Search Screen from the "Omic Browse" Database

The OmicBrowse Database facilitates integrated searches for and visualization of the wide range of information accumulated to date through genomics research.

The Integrative Omics Research Team is constantly engaged in bioinformatics research and development for integrated data analysis. Furthermore, through collaborative research efforts, they are analyzing the vast amount of experimental lab results and data that is essential for performing integrated data analysis.
A concrete example of the team's work is the analysis of ENU mutant mice for the GSC animal group, where the use of the disease gene prediction system (PosMed) developed by the team optimized the search for genes responsible for disease, and brought over 50 successful results. This system is also available through the Internet, and is used frequently in Japan and overseas. Furthermore, the statistical analysis technology and the system for analysis of integrated data developed by the team is now used as a fundamental technology to process the vast amount of experimental data produced from DNA chips used by the Plant Science Center, Research Center for Allergy and Immunology, and the BioResource Center at RIKEN. Furthermore, the integrated search technology developed by the team is employed as the core technology in the RIKEN integrated database, thus it also functions for RIKEN as data acquisition and data release infrastructure. Furthermore, the software for integration of genome annotation, "OmicBrowse", was released as open-source software, allowing researchers across the world to bring together hundreds of different types of databases into one integrated whole.

Understanding Systems of Life: the Ultimate Goal

In the future, we will conduct interdisciplinary research between high-performance computing, informational/statistical science, and life sciences, aiming for an understanding of systems of life. In order to achieve this goal, we promote advanced computational life science utilizing the power of future high-performance computers and large-scale cyclopedic data sets, and we also develop systems biology research under tight collaboration between experiments and computation. In the future, we intend to enhance our activities to include developmental systems biology, metagenomics as a systems analysis at the environmental level, and synthetic biology based on a systems approach.
In addition, a framework for handling various kinds of integrated data will become ever more essential as the quantification of biology continues at breakneck pace, so we are positively involved with the relevant information technology and services for creating this framework.