University of Tokyo's RCAST and Fujitsu Collaborate in Race to Develop New IT-based Drug Discovery Technology

Tokyo, June 10, 2011

The University of Tokyo's Research Center for Advanced Science and Technology (RCAST) and Fujitsu today announced their collaboration to develop new technology to efficiently design small molecule drug candidates for cancer and other indications.

The joint research project will employ Fujitsu's small molecule design and evaluation technology⁽¹⁾—a computer-based system that generates novel chemical structures and predicts their efficacy as drug candidates. The result will be enhanced capabilities needed to design the small molecule compounds that will become drug candidates, made possible with a high level of precision on par with actual biochemical assays, which is difficult to accomplish with conventional methods. Another benefit is that it will become possible to conduct simulations that evaluate proteins, as well as their behavior, with regard to the cause of specific disease indications.

Fujitsu and RCAST aim to develop new IT-based drug discovery technology within three years by combining Fujitsu's technology with RCAST's biological data, such as its research on proteins that are believed to cause specific disease indications, as well as its chemical experiment technology and know-how. In so doing, they aim to make possible, for a variety of disease indications, the efficient design of small molecules with completely novel chemical structures that can selectively target proteins underlying specific disease indications. The goal is for the results of this joint research to promote the systematic development of new drugs with high efficacy while contributing to improved quality and exceptionally cost-effective medical care.

New IT-Based Drug Discovery

The human body is comprised of many different proteins, and it is the interactions among these proteins that sustain human life. When the balance among these proteins is disrupted, it triggers the development of disease. Drugs are designed to control the activity of proteins that are believed to cause specific diseases, thereby restoring the body's normal balance. Accordingly, the process of drug discovery first requires an identification of the proteins that cause a specific disease, followed by the design and synthesis of compounds that modify the target proteins, and then verification of the results (see Figure 1).

With this new IT-based drug discovery method, simulation technology is used at the compound design stage to virtually design, on a computer, a compound structure that effectively interacts with the structure of the identified target protein, enabling the swift and inexpensive generation of highly effective new compounds. With conventional technology, it was difficult to simulate the environment of a patient's body after a drug has been administered. However, Fujitsu's technology and advances in computing performance have now made such simulations possible. The new IT-based drug discovery technology is therefore anticipated to become a next-generation drug discovery method for generating highly effective new compounds in a short period of time and at a low cost.

Small Molecule Compounds

Small molecule compounds are compounds that have a low molecular weight that have the potential to become drug candidates. Most marketed drugs are small molecules that can be produced economically in large quantities through chemical synthesis. Small molecules have a variety of different chemical structures, which means the introduction of new IT-based drug discovery technology shows great promise in supporting the development of groundbreaking drugs that selectively target proteins underlying specific disease indications.

Background to the Current Research Collaboration

Last year RCAST installed a cluster supercomputer (3,600 CPU core) comprised of Fujitsu's PRIMERGY BX922 S2 blade servers. RCAST has been using this equipment in line with its research goals of generating antibody drug therapies for relapsed, metastatic cancer indications. Through supercomputer simulations of molecular dynamics⁽²⁾, RCAST is conducting research to systematically design an antibody drug, and has characterized biological dynamics at a molecular level to an extent that was unattainable through previous methods.

For its part, Fujitsu has been developing and marketing computational chemistry software for over 20 years, and the company has world-class technology and talent covering the entire computational chemistry field, from theory to practical applications. In 2004, to accelerate its initiatives in the biopharma field, Fujitsu established its Bio-IT Business Development Unit and started its activities in the field of IT-based drug discovery. Through in-house research as well as research collaborations with companies, universities, and research laboratories around the world, it has developed highly sophisticated IT-based drug discovery technology.

When RCAST decided to begin a new R&D initiative designed to generate small molecule drugs, it recognized the value of Fujitsu's small molecule drug design and evaluation technology, which led to the current research collaboration.

Technological Challenges

In the traditional drug discovery process, chemists would rely on their experience and knowledge. They would select from among a group of small molecule drug candidate compounds those most likely to be effective, and then they would perform a series of biochemical studies to develop actual drugs. Even with IT-based drug discovery, it is inherently difficult to conduct validation studies on all small molecule compounds that are potential drug candidates, of which there are said to be more than 10 to the 30th power. Computer-based screening is therefore performed prior to the search. It would be ideal if the screening method could come close to replicating the biological environment of the human body, but the current screening methods used are not designed to closely replicate the environment of the human body in order to efficiently screen very large volumes of molecules. This raises questions about the reliability of results. In addition, because chemical structures generated from refinements to existing compounds tend to be very similar to the existing compounds, it has been difficult to quickly generate highly effective drugs having entirely novel chemical structures.

Research Collaboration

The current research collaboration is focused on designing small molecule compounds. It will use RCAST's research on proteins believed to be the underlying cause of specific disease indications, and combine it with Fujitsu's OPMF⁽³⁾ small molecule design software, which enables drug-efficacy prediction and screening with a degree of precision equivalent to actual clinical trials, and the company's MAPLE CAFEE⁽⁴⁾ accurate binding activity prediction software to perform simulations run on the supercomputer developed by Fujitsu. The computer-designed compounds designed will then be evaluated by RCAST through a series of biochemical assays (figure 2).

RCAST and Fujitsu, with the latest research results and leading-edge technology, respectively, have set as their goals the development of technology that will become a foundation for the development of drugs targeting as yet untreated disease fields, as well as the generation of small molecule compounds within three years.

Aspirations of the Research Collaboration

If this joint research project is able to generate small molecule drug candidates for oncology, it may be possible to develop highly-effective treatments for cancer. Moreover, the development and application of new IT-based drug discovery technology has the potential, in the future, to generate new treatments for a variety of indications.

Overview of the Research Collaboration

• Timeframe:
  June 2011 to March 2014
• Structure:
  RCAST: Dr. Tatsuhiko Kodama and 5 other researchers
  Fujitsu: Shunji Matsumoto, General Manager of the In-Silico Drug Discovery Research Division, Bio-IT Business Development Unit, and 10 other researchers
• Location:
  A Fujitsu office will be established inside the University of Tokyo Research Center for Advanced Science and Technology (RCAST)