An AI Startup is trying to find an effective treatment for COVID-19

The U.S based Insilico Medicine is utilizing machine learning models to identify molecules to devise a treatment

Faisal Khan
Feb 13 · 3 min read

At the recently concluded meeting of the World Health Organization (WHO), the recent virus outbreak out of China was renamed as the “COVID-19” from Coronavirus to avoid stigmatizing entire regions or ethnic groups as we saw historically in cases like the “Middle East Respiratory Syndrome” or “Spanish flu.” In the newly coined word— “CO” stands for “corona”, “VI” for “virus” and “D” for “disease”, while “19” marks the year it was first identified.

The global health agency was cautiously optimistic that although new cases were slowing down, the disease still represented a significant threat and appealed for speeding up research into finds viable drugs and/or vaccines for the outbreak. The most recent numbers at the time of publishing this article were as follows: Confirmed cases 60,179 — Deaths 1,365 and Recoveries 5,924. It is pretty evident that the threat hasn’t subsided.

Meanwhile, efforts seem to be gathering pace to find a vaccination or cure for the disease which has rocked the World. A U.S based startup, Insilico Medicine has announced that is has used Artificial intelligence to identify molecules that can expedite the formation of treatment for COVID-19. The whole idea behind using advanced biomedical technology is to find a cure for the deadly disease in a couple of weeks rather than months or even years that would otherwise take.

Insilico set out by first studying the long list of possible targets for a 2019-nCoV treatment published by the Beijing-based Global Health Drug Discovery Institute. Starting on Jan. 31, it started utilizing 28 different machine learning models to find the relevant molecules to devise a cure.

It took the company’s A.I.-based system only four days to identify thousands of new molecules. Of the tons of molecules identified, Insilico would test 100 of the most promising candidates. The complete list of the rest of the molecular structures would be published for the other researchers if they want to use it.

Some of these machine learning techniques involved employing generative adversarial networks (or GANs). Just like GANs use in creating deepfakes, they replicated new molecules that have the right structure for effective drugs. The company further used a filtering process on these GAN produced molecules to further shortlist the ones that have high “drug-like” properties. This filtering also helped in phasing out the molecules that already exist.

However, Insilico is not the only company making use of AI to find a cure for COVID-19. Researchers from Michigan State University have also published a paper on using machine learning techniques to create new drug candidates. Another U.S.-based biotechnology firm Gilead, in collaboration with a Beijing hospital, has already started human testing of an existing antiviral drug, remdesivir, in Wuhan.

Complete research was published on the repository Research Gate and on the company’s website. Insilico has invited peer review by other researchers to accelerate the process of finding a cure.


Empowering you with Technical, Scientific & Financial knowledge

Faisal Khan

Written by

Content Specialist in Cryptocurrencies | Blockchain | Financial Markets | Technology | Future | Science | Space


Empowering you with Technical, Scientific & Financial knowledge

More From Medium

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade