A research group led by Professor Akaji of Kyoto Pharmaceutical University has developed a drug discovery research method incorporating AI technology through joint research with Preferred Networks Co., Ltd. (PFN).
In addition, using this technology, we have discovered a lead compound *1 for a drug to treat COVID-19, a new type of coronavirus infection .
Table of contents
- The drug discovery industry takes more than 10 years and costs tens of billions of yen
- Creating lead compounds for new coronavirus therapeutics using drug discovery AI
The drug discovery industry takes more than 10 years and costs tens of billions of yen
In general, drug discovery takes more than 10 years from basic research to manufacturing and marketing, and the cost is said to be more than tens of billions of yen. Also, the success rate is extremely low, less than 1 in 20,000.
In the drug discovery industry, the enormous amount of time and money required for development has become a major issue.
Under such circumstances, AI drug discovery, which incorporates AI into the drug discovery process, has introduced deep learning and large-scale computing resources to the field of drug discovery research , which has so far relied heavily on the knowledge of researchers. I’m here.
As a result, it is expected that it will be possible to drastically cut the drug development process such as searching for candidate substances and molecular design, and to propose new molecular structures that humans cannot conceive .
Image of a drug (yellow) that binds to the main protease (purple) of the new coronavirus and inhibits its growth
(for intellectual property protection, a different substance than the one discovered this time is used as a reference example)
Creating lead compounds for new coronavirus therapeutics using drug discovery AI
Kyoto Pharmaceutical University has been conducting basic research aimed at developing therapeutic drugs for severe acute respiratory syndrome (SARS), and has been developing ” peptide compounds ” and ” non-peptide cyclic compounds ” . was creating These two compounds work by inhibiting enzyme functions that are essential for the replication of the SARS-causing virus.
Through this basic research and AI drug discovery developed by PFN, we have created a promising lead compound that can inhibit the enzyme (main protease *2 ) essential for the proliferation of the new coronavirus (SARS-CoV-2) .
In this research, 13 of the multiple compounds proposed by PFN’s AI drug discovery and private supercomputer MN-2 were synthesized at Kyoto Pharmaceutical University. As a result of evaluating their activity, about half of the 7 compounds were able to inhibit the activity of the main protease of the new coronavirus .
The fusion of the results of basic research at the university and AI drug discovery has given the policy of compound design a high degree of feasibility. It also leads to evaluation of the rationality of computer-designed compounds as drugs.
The accumulation of basic research and the fusion of the latest AI drug discovery will contribute to new drug development methods in the future.
