Chances Are as Slim as One in 30,000. In Silico Drug Discovery Has the Possibility to Break Through the Reality
Drug discovery refers to the development of new drugs. This requires an immense amount of time, effort and costs. Some say that even a medicine that is readily available at pharmacies and drug stores requires 15 years or more and a cost of hundreds of millions of dollars for research and development. Even worse, with this amount of effort, the chances of being used as an actual medicine are no higher than approximately one in 30,000 (according to a survey conducted by the Japan Pharmaceutical Manufacturers Association).
It is expected that in silico drug discovery will help solve these issues. This is an approach to creating new drugs as fast and efficiently as possible by improving development time, efficiency and accuracy with IT.
Create Unknown Compounds and Use Them to Facilitate the Treatment of Irremediable Diseases
Higher efficiency in new drug development is not the only advantage of in silico drug discovery. Another key role is to create new compounds that have never existed in the world.
Compounds constitute drugs. The human body is mostly comprised of protein. A compound absorbed into the body acts on the protein that causes a disease, thereby serving as a drug. For this reason, drug discovery research starts with identifying the protein that causes the disease and finding a compound that acts on the protein.
Conventional drug discovery involves trial-and-error experiments to discover compounds that may serve as drugs using existing compounds. However, there are an immense number of protein-compound combinations. In other words, an uncountable number of experiments can be conducted, which is a key factor that has kept the chance of successful drug discovery low.
Conversely, with in silico drug discovery, the characteristics of a protein, including the shape, can be used as a basis to create a new compound that acts on the protein. The creation of unknown compounds can lead the way to producing drugs effective against diseases for which no drugs have been available and difficult-to-treat diseases.
Compounds with Specific Shapes Cannot Serve as Drugs
What does it mean that a compound acts on a protein? This has a great deal to do with the physical structures of proteins and compounds.
There are many types of proteins, each of which has a unique shape with dips, holes, projections, etc. To act on a target protein, a compound must have a shape that fits into the shape of the protein.
Assume that there is a protein with an L-shaped hole in it. A compound that acts on it must have an L-shaped projection. However, an L-shaped projection cannot go into the entrance of the hole.
Fortunately, both compounds and proteins can change their shapes. In particular, predicting how compounds referred to as small molecules change their shapes is more difficult than proteins consisting of 20 types of amino acids because it is said that there are 1060 small molecules. Here, the focus is more on how compounds change shapes.
What if a compound could change its projection into an I shape and L shape? If the compound changes the projection into a straight shape before intruding into the protein's hole and then changes the projection into an L shape after intrusion, it could bind to the protein.
New Technology to Simulate How Compounds Change Their Shapes
Conventional drug discovery involves repeated experiments reliant on intuition and experience in order to discover such protein-compound combinations. For this reason, the success ratio is extremely low and the majority of such experiments failed.
Even with in silico drug discovery and supercomputers, simulating how compounds change their shapes was difficult. Predicting how shapes change was very difficult, which prevented the practical use of in silico drug discovery in drug discovery research.
Fujitsu developed a new molecule simulation technology that helps predict such shape changes. This makes it possible to simulate compound shape changes accurately enough with currently available computer capacity.
The leveraging of molecular simulation technology will facilitate the efficient discovery of compounds that act on diseases. This will surely lead to reducing the time required for drug discovery and development.
Successful Creation of Cancer-Targeting Compounds Using Advanced Technology and Co-Creation
The in silico drug discovery initiative has also resulted in the successful creation of a new active compound that targets cancer cells. This achievement is attributable to co-creation with the University of Tokyo's Research Center for Advanced Science and Technology (RCAST) and Kowa Company, Ltd. in addition to Fujitsu.
Fujitsu designed on computers many new compounds that should act on a protein thought to cause cancer. We proceeded to conduct an experiment that involved the synthesis of eight compounds likely to fit into the shape of the protein and act on it stably, which showed that one of them has a cancer-blocking effect. Compared to the conventional drug discovery method, it is safe to say that the ratio of one eighth, or 12.5%, is very high.
Fujitsu set forth in silico drug discovery in 2004. When we started to work on co-creation with pharmaceutical companies after developing the technology, pharmaceutical companies commented that new compounds that we created with IT did not appear to be drugs (in other words, their chemical structural formulas did not seem to be those for pharmaceutical compounds). However, our continued research and development has made it much more possible to create drug candidate compounds that have never seen.
There are still many diseases that cannot easily be treated such as serious diseases, incurable diseases and lifestyle diseases. Going forward, Fujitsu will strive to work on in silico drug discovery through co-creation with other companies in addition to developing unique technologies so that we can help save people suffering from such diseases.
Nozomu Kamiya, Ph.D.
Director, Bio-IT R&D Office,
Research and Development Division,
Healthcare Solutions Unit II,
Azuma Matsuura, Eng.D.
Business Acceleration Group,
Digital Annealer Project,
Fujitsu Laboratories Ltd.