Laying the Groundwork for In Silico Drug Discovery with the K computer to Increase the Pace of New Drug Development


Laying the Groundwork for In Silico Drug Discovery with the K computer to Increase the Pace of New Drug Development

Osaka has long been the home to many pharmaceutical companies in japan. The NPO Biogrid Center Kansai, located in Grand Front Osaka (opened in 2013), partners with universities and pharmaceutical companies to lay the groundwork for in silico drug discovery using the K computer. In silico means leveraging new IT-based techniques for drug discovery in addition to traditional experiment-based techniques.

"We originally began with the mission of match-making supercomputers and biotech, then in 2011 we learned that construction of the K computer would be close to an end in Kobe. So we wondered if industrial use of supercomputers for pharmaceutical industry such as drug discovery would be possible. We began efforts to leverage the K computer for industry right around the time that it became available for shared use."

Biogrid Center Kansai helps establish links between pharmaceutical companies and academia to transfer IT-based drug discovery techniques developed by Professor Okuno at Kyoto University. A total of 11 pharmaceutical companies joined the project to forge an infrastructure for in silico drug discovery using the K computer which involves discussions breaking down barriers between corporations on a regular basis.

"Most people cannot imagine using computers to develop drugs. But the K computer enables unprecedented precision in calculations, so we are developing methods to efficiently create drugs that will be effective against the proteins causing-diseases."

Professor Okuno’s research involves using the K computer to rapidly calculate interactions between about 600 disease-related proteins and about 30 million chemical compounds. Predictive accuracy is enhanced even further with rigorous simulations. All pharmaceutical companies active in the project share these results, which help them to form the basis for developing the next generation drugs in their own company.

"It may be surprising to learn that predictive accuracy of conventional computations used to be used in drug discovery, both by us and pharmaceutical companies in ordinary computational environment, is only about 5%. Our aim in using the K computer is to dramatically improve this figure. My goal, or dream is to use the K computer or even the next generation supercomputer beyond that to achieve predictive accuracy of 70% or higher. I have great hopes that the world of IT-based drug discovery will soon make this possible."

In future, if the supercomputer much more powerful than the K computer becomes available, it is expected that a proper chemical compound will be automatically selected as a candidate if the protein causing a disease is input into the computer.

To contact about this post.

blog comments powered by Disqus



  1. 08
  2. 08
  3. 08
  4. 08
  5. 08


  • Today
  • This Week
  • This Month
  • All Time