Covid-19 Therapies Designed with Virtual Reality, A.I.

Image of SARS-CoV-2 protease (Nanome Inc.)

22 May 2020. Two companies are using deep learning and virtual reality to design treatments for Covid-19 infections, attacking a different target than most drugs being developed. Insilico Medicine in Hong Kong that designs drugs with artificial intelligence and Nanome Inc., a virtual reality software developer in San Diego outline their process in an unedited paper posted 19 May on the pre-review site ChemRxiv.

In the paper, Insilico Medicine and Nanome describe 10 small molecule, or low molecular weight, therapies for Covid-19 infections addressing the protease in the virus, a different target than most treatments now in development. So far, new Covid-19 drugs focus mainly on the spike or S protein on the viral surface that penetrates cells and begins the infection process. The therapies designed from scratch by Insilico Medicine and Nanome, however focus on the protease, a protein the virus needs for replication. Many antiretroviral drugs treating HIV infections, for example, are protease inhibitors that block the virus’s replication.

Insilico Medicine discovers drugs with what the company calls generative chemistry that uses machine learning algorithms to design compounds addressing molecular targets with needed properties. This approach, says the company, allows for the structure of the target and desired properties to dictate the chemistry of new treatments.

Using massive databases for training algorithms with deep machine learning, Insilico says it can quickly design therapeutic candidates addressing specified targets and with specified properties. As reported by Science & Enterprise in September 2019, researchers from Insilico and colleagues designed new small molecule drugs that block an enzyme associated with cirrhosis, or scar tissue in the liver, ready for validation and lab animal tests in 21 days.

Insilico says it began in late January 2020 to discover new treatments for Covid-19 infections. The company says it focuses on blocking the protease, building on the experience of some treatments for SARS and MERS infections also designed as protease inhibitors. The design as well specified weaker, non-covalent bonds with the protease.

“Covalent inhibitors,” says Insilico’s CEO Alex Zhavoronkov in a company statement, “form a strong bond with a protein, but they aren’t very selective and usually cause severe side effects. To date, the most potent inhibitors of the SARS-CoV-2 protease are covalent. We set out to generate non-covalent drug candidates due to their higher levels of safety and efficacy.”

Insilico is partnering with Nanome to refine the proposed Covid-19 therapeutics. Nanome uses virtual reality or V.R. to visualize, measure, and analyze drug molecules as 3-D structural images. The V.R. sessions are also designed for online collaboration in 3-D, allowing for iterative testing of proposed molecules in real time and close inspection of results.

Zhavoronkov adds that “it is important for medicinal chemists to look at these molecules closely before placing a billion-dollar, life-or-death wager. V.R. enables medicinal chemists to do this while unleashing creativity and encouraging collaboration across both disciplinary and physical borders.”

This 10-minute video shows Nanome and Insilico scientists collaborating on proposed Covid-19 treatments addressing the SARS-CoV-2 virus’s protease. Nanome is also offering its services to the Covid Moonshot, an international group of scientists addressing the SARS-CoV-2 protease.

More from Science & Enterprise:

• Symptom App, Algorithm Predict Covid-19 Infections
• Pharmas License A.I.-Aided Gene Therapies
• Biotech Partners for A.I. Drug Discovery in $1B Deal
• A.I. Discovery Companies to Find Rare Disease Drugs
• Viruses Altered to Boost Gene Therapy Delivery

*     *     *