次世代スパコン「ポスト京」が拓くバイオスーパーコンピューティング Biosupercomputing opened by next-generation supercomputer post-K
金沢大学 角間キャンパス 自然科学本館 G会場（レクチャーホール）
Supercomputers continue to grow in scale and capability after the K computer in the world. In Japan, the development of the post K computer has been launched, and the huge computation power of the post K computer will be expected to push the boundary further in computational life sciences such as biomolecular simulations, genomic analysis, drug discovery, and health care. In this symposium, the future perspectives of biophysical supercomputing in the post-K-computer era will be discussed.
Unraveling cancer heterogeneity with supercomputer
Satoru Miyano（The University of Tokyo）
We present highly parallel software applications developed in the project“HPCI Strategic Programs for Innovative Research Field 1“Supercomputational Life Science” (2011-2015) and the Grand ChallengeProject for Life Science “Next-Generation Integrated Simulation of Living Matters (2006-2012). The first series of applications include various genenetwork estimation software applications and the second series include a series of software applications and pipelines for cancer genome analysis (exome, whole genome, RNA sequence). By using these software applications, we are currently making challenges to understand cancer heterogeneity on network-level from cancer big data.
Hierarchical Integrated Simulations of Circulatory System
Shu Takagi（The University of Tokyo）
There are several stages of the simulations for cardiovascular system. We have been developed the software for entire vascular network model using the concept of 0D-1D-3D coupling. We have been also developing the multiscale thrombosis simulator. Recently, we have started the 3-D simulations for micro-circulation systems. So far, although these simulators are our own made, that is, developed with our colleagues, they are not unified for hierarchical integrated simulations of circulatory system. In the present talk, these simulators are briefly introduced and toward their hierarchical integration for the post “K-computer”, future direction is discussed.
Biomechanical simulation integrated with clinical measurements toward personalized medicine
Shigeo Wada（Graduate School of Engineering Science, Osaka University）
The development of biomedical measurements such as MRI and X-ray CT has led to great progress in clinical medicine. On the other hand, recent advances in computational mechanics have enabled to simulate biological phenomena at various scales from cells to organs, bridging the distance between mechanics and biology. Thus, it is expected that the integration of the computational approach with the clinical data assists in individual diagnosis and treatment. However, the computational approaches have not been generally accepted in the clinical medicine where the biomedical measurement is emphasized as the evidence. This presentation describes the problems to be solved toward the personalized medicine.
Innovative drug discovery infrastructure through functional control of biomolecular systems by using post ‘K’ supercomputer
（1RIKEN Quantitative Biology Center, 2Graduate School of Medicine, Kyoto University）
In the end of last year, developing of a next generation supercomputer, called post ‘K’ had been started. Following this project, we are also developing application and software for drug discovery and development by using post ‘K’. The application project aims to develop ultra-high speed molecular dynamics simulations to achieve not only capturing long-time dynamics of target molecules in the time range of millisecond, but also controlling supra-molecular systems constituting of many biomolecules including factors that cause side-effects. We hope that post ‘K’ will accelerate discovery and design of safer and more highly effective drugs.
Large-Scale Molecular Simulation of Viruses: Multi-scale Molecular Modeling Approach
Wataru Shinoda, Kazushi Fujimoto, Yoshimichi Andoh, Susumu Okazaki
（Grad. Sch. Eng., Nagoya Univ.）
We illustrate here our recent work on the molecular modeling and simulations of large self-assembled macromolecular systems including virus capsid. Previous all-atom molecular dynamics simulations elucidated a special environment within a virus capsid, which motivated us a further investigation of molecular adsorption and entrance to the capsid in the context of drug design. To characterize the molecular mechanism, an efficient way to evaluate the free energy surface is desired. We here try a multi-scale molecular modeling approach to overcome this problem. We would like to discuss also a possible future work using high-performance supercomputers.
Conformational Dynamics of Multidrug Efflux Transporter AcrB
Mitsunori Ikeguchi（Grad. Sch. Med. Life Sci., Yokohama City Univ.）
Conformational dynamics of target proteins are crucially important in drug development especially for flexible proteins. Controlling the conformational dynamics of target proteins is one of major goals for HPC drug development using the post K supercomputer. In this talk, I present a computational study on conformational dynamics of the multidrug efflux transporter AcrB, which undergoes large conformational changes during drug transports. AcrB utilizes the concentration gradient of protons across the inner membrane of bacteria as an energy source. Coupling between two conformational changes in transmembrane and periplasm domains is responsible for dynamical functions of AcrB.