Please view the main text area of the page by skipping the main menu.

Japanese supercomputer finds 30 existing drugs potentially effective to treat COVID-19

This image provided by the Riken research institute and Kyoto University professor Yasushi Okuno shows how antiparasitic drug Niclosamide, in pink, and SARS-CoV-2 proteins bind together in a simulation carried out by supercomputer Fugaku.
Supercomputer Fugaku, developed jointly by the Riken research institute and Fujitsu Ltd., is seen in this photo taken in Kobe's Chuo Ward on June 23, 2020. (Mainichi/Tatsuya Onishi)

TOKYO -- Analysis by Japanese supercomputer Fugaku, which was recently named the world's fastest, has found some 30 existing medications that could be effective COVID-19 treatments, Kyoto University professor Yasushi Okuno announced on July 3.

    According to the supercomputer's assessment, the candidate drug with the most potential was not known to be effective for coronavirus treatment. Professor Okuno, a specialist in computational science for pharmaceutical development, says a Japanese firm holds the patent for this existing drug. He hopes to launch clinical research and a drug trial after negotiating with the manufacturer.

    Okuno and a team of researchers ran simulations using Fugaku to see how 2,128 existing drugs, including anticancer agents and common cold medicine, bind to proteins unique to SARS-CoV-2 -- the novel coronavirus' scientific name -- and looked into how the drugs worked in the body on the molecular level. It appeared that the longer the agent stayed locked inside the keyhole-like spaces of the proteins, the stronger the binding and thus more likely the drug would be effective. The research team ranked the medications based on how long the agents were connected to the proteins.

    The 12 drugs currently being tested overseas as potential COVID-19 treatment all ranked high in Fugaku's simulations, including antiparasitic drug Niclosamide, which came in second place. The medication that connected with the SARS-CoV-2 proteins the longest was Japanese, but Okuno says he can't name the drug due to potential patent infringement. Many other medications, whose efficacy as coronavirus treatments had been unknown, also ranked high in the simulations.

    The protein targeted in their research was the main protease of SARS-CoV-2. This enzyme works to replicate viruses in a human cell.

    The team plans to conduct simulations on three other types of SARS-CoV-2 proteins and compile data by the end of the summer. It took Fugaku 10 days to run calculations on 2,128 drugs, but it would have taken at least a year if it was done by Fugaku's predecessor, the Riken research institute's K supercomputer, according to the team.

    (Japanese original by Tomohito Ikeda, Science & Environment News Department)

    Also in The Mainichi

    The Mainichi on social media