“We expect COVID-19 to become a persistent human virus like flu,” says professor Nikolai Petrovsky in an interview conducted by Anna Zagórna

In January we wrote about the researchers from Flinders University, Adelaide, who were using artificial intelligence to work on a cure for 2019-nCoV, aka the Wuhan coronavirus. Their algorithms and simulations carried out by supercomputers in July 2018 have developed a vaccine for seasonal flu. It is now being tested in the United States. The team responsible for the research is led by professor Nikolai Petrovsky. As soon as the coronavirus proved to be a global issue, he decided to use supercomputers and artificial intelligence to create a vaccine.

Anna Zagórna: What stage is your team currently at?

Nikolai Petrovsky*: We would like to test our vaccine on animals this week. Before we conduct trials on humans, we want to make sure that it is safe and effective. Other companies start their trials directly on humans, but we think this is unwise. We hope our vaccine will be tested on human volunteers in about three to five months.

The supercomputer and cloud computing played a major role in the creation of the drug.

To identify potential medicines that could stop COVID-19 we used computer screening methods. We have been also using computer simulations to create COVID-19 models, which helps us a lot with development of vaccines. It makes the whole research process much shorter. With that solution, we can save several years.

You were scheduled to come to Kraków in March but the epidemic interfered with your plans. As far as I know, you and your team are now working on the vaccine from home.

I’ve been in the United States since January and my research team is located mainly in Australia, although some of its members are now scattered around the US. All communication regarding development of the vaccine and the cure for COVID-19 has been limited to internet, Skype, Zoom etc.

We hope that artificial intelligence will help us to identify an already existing medicine that might be used once again to treat people infected with COVID-19

Nikolai Petrovsky

In our research work we also use cloud-based supercomputers, which are accessible via the internet. Both the results of our research and our ideas are also uploaded to LinkedIn. We would like to publish our findings regarding COVID-19 on prepress servers [Editor’s note: before or during their review] in order to make them immediately available.

On Monday the White House Office of Science and Technology Policy called the scientists to use the artificial intelligence technology to analyze 29 thousand scientific articles in search for the answers to key questions regarding the coronavirus. How does your team use artificial intelligence?

We hope that artificial intelligence will help us to identify an already existing medicine that might be used once again to treat people infected with COVID-19. It is much faster than developing a completely new drug from scratch. We don’t have that much time and we can’t wait.

Is there anything else that would speed up your work on the vaccine?

Only money. Currently, we are entirely dependent on our internal resources.

The trials you plan to conduct on humans will be very difficult as people’s reactions to the disease are different depending on their age. The elderly and the sickly are much less resilient. What will the tests look like?

First, we will vaccinate healthy adults. Then, we will proceed with the trials on the elderly and the youth. This is the safest approach. We would administer every discovered and tested antivirus drug to every person suffering from a severe condition of COVID-19, irrespective of their age. However, it must be noted that it probably won’t be possible to fully wipe out the coronavirus from our environment.

Do you think it will stay with us forever?

Yes. We expect COVID-19 to become a persistent human virus, just like flu, unless we succeed in developing a highly-effective vaccine.

Is the scientific community collaborating within the scope of the research? Do scientists make the data available to each other?

Many Chinese researchers and clinicians have done an excellent job by putting out the first COVID-19 genome sequence, first clinical case reports and first results of trials on animals. The rest of the world, including us, used the results of the Chinese to conduct our own research.


*Nikolai Petrovsky is a professor of medicine at Flinders University and the director of the Department of Diabetes and Endocrinology at Flinders Medical Center, a hospital clinician, and the vice-president and secretary-general of the International Immunomics Society. He has developed vaccines against influenza, hepatitis B, sting allergy, malaria, Japanese encephalitis and rabies. He authored over 100 papers and chapters and won prestigious awards, including the AMP Innovation Award at the Telstra Business Awards 2009 and the 2010 Ernst & Young Entrepreneur of the Year award.


Skipping the animal testing stage

The first human trial of vaccine against the coronavirus has started in the US. The vaccine developed by the researchers from the drug company Moderna Therapeutics called mRNA-1273 contains neither killed nor weakened virus; specialists in genetic engineering recreated a part of RNA responsible for the production of a protein allowing the virus to penetrate human cells. Moderna Therapeutics has been using artificial intelligence to considerably speed up the machine learning cycle, ensuring critical insight of the data regarding trials and production. However, the company keeps its method secret after having surpassed other companies in terms of their efforts to develop vaccines.

The experiment will involve 45 healthy adult volunteers aged 18 to 55. Each of them will receive 100 dollars for each of the in person study visits, or a total of 1,100 dollars.

The first dose was administered to four inhabitants of Seatle in Kaiser Permanente Washington Health Research Institute. Each of the participants will receive two vaccine doses 28 days apart. This is an unusual way of running vaccine tests. Usually regulatory authorities obligate producers to prove that a product is safe before it is distributed to or used on people. Researchers almost always use lab animals to check if a new drug is effective before exposing the volunteers to a risk.

According to the Moderna’s recipe mRNA is synthetic and programmed to make us capable of producing some proteins that are similar to the coronavirus. As soon as they appear, our bodies will learn how to identify and block real viruses. The biggest advantage of that method is that it is fast. The vaccine was developed within less than one week after the new virus had been discovered and was ready on 13 January. On 24 February the team sent the vials from the facility in Norwood, Massachusetts, to the National Institute of Allergy and Infectious Diseases in Bethesda to schedule clinical trials to assess the safety of the vaccine.

Skip to content