AI VIVO identifies list of 31 drugs that show potential for Covid-19 treatment
A shortlist of 31 candidate drugs that are already approved or in late clinical stage trials has been identified in the search for potential Covid-19 treatments by Cambridge company AI VIVO.
It used its artificial intelligence-driven phenotypic approach, assessing observable characteristics to rank 90,000 candidate compounds in just 15 days.
The company, based at the Bio-Innovation Centre on Cambridge Science Park, used real samples from Covid-19 infected cells to build its model of the impact of the virus on the lungs.
Its prediction engine then delivered its list of 31 promising candidates, which represent 0.05 per cent of those studied.
One of the company’s lead investors is serial entrepreneur David Cleevely, who is also a government adviser.
He said: “We are already talking to a number of government agencies and pharmaceutical organisations about testing the top ranked drugs on our list which could prove to be even more significant and make a real difference in the fight against Covid-19.
“Now that we have the crucial data, it’s essential that we get on with testing these drugs as quickly as possible. And, to that end, we’d be delighted to hear from anyone interested in partnering or collaborating with us.”
Encouragingly, among the top 31 are five drugs that have already entered clinical trials for Covid-19:
- Chloroquine, an anti-malarial drug;
- Dexamethasone, a commonly used corticosteroid;
- Sirolimus, an immuno-suppressant drug;
- Tacrolimus, an immuno-suppressant drug;
- Thalidomide, a cancer treatment drug.
Chloroquine has attracted much attention after being mentioned by US President Donald Trump and has also been the subject of a promising small trial in France.
AI VIVO’s system suggested how the drug’s efficacy could be improved and side effects moderated by combining it with other compounds in the top 31.
Founder and CEO Dr Peyman Gifani said: “We are a team of molecular biologists, system pharmacologists, control engineers and machine learning experts.
“We believe identifying phenotypic changes in disease states is key to finding effective therapies.
“There is little known about the Covid-19 virus itself, its effect on the host tissue and potential targets for discovery research. This limits our knowledge about the best targets for Covid-19 and means that AI VIVO's phenotypic approach could provide the most powerful, logical and rapid pathway to Covid-19 drug discovery.
“We are now in the process of extending our analysis to wider combinations of treatment strategies. This will lead to the identification of combinations of top ranked drugs that, together, will be better than any single drug on its own.”
AI VIVO said for the Phase II and III candidates it shortlisted, it can help pharma or biotech proprietors speed up their process and identify effective drug combinations for Covid-19 treatments.
The company wants to hear from potential partners. It can be contacted on covid19@aivivo.co.
Dr Peyman Gifani discusses AI VIVO’s approach
Everyone is hoping for a Covid-19 cure, but the path to finding effective drugs for any disease is slow and difficult. Drug compounds must be painstakingly identified and many phases of testing are required to make sure they are safe and actually do the job.
At AI VIVO, a maverick biotech company based in Cambridge, we have jumped into the race for a cure and already offered new hope, predicting 31 new drugs that are highly likely to be effective in treating Covid-19 disease.
When a virus like Covid-19 attacks the body, it targets any points of vulnerability in the human body’s defence system that it can find.
Scientists fight back by trying to find the virus’s own vulnerabilities to target with drugs, drugs that either shut Covid-19’s access to human cells or disable its core functions.
In the race for a Covid-19 cure, scientists and their computers scramble to find targets faster than the virus can. These scientists use their experience to make educated guesses about which drugs might work. With infection rates rising around the world, we’ve so far been losing the race.
But AI VIVO has approached the challenge differently. While most of our pharma and biotech colleagues are searching for molecular targets, our team are looking at a much larger picture: how lung cells behave when faced with Covid-19.
When the virus attacks a person’s lungs, the lung cells are forced to react, changing how they copy genes and build proteins, in a desperate attempt to fight the virus off. This chain of reactions is what we call the ‘Covid-19 disease’, producing typical symptoms like coughing and fever.
The AI VIVO team has built an artificial intelligence system that tracks the reactions of healthy cells as they transform into diseased cells. The system tracks genetic patterns over the whole human genome (20,000 genes) contained within lung cells. The system then looks for ‘backdoor access’ to Covid-19 and identifies a collection of drugs that will help diseased cells become healthy again.
Until a few years ago, scientists didn’t have the computational power or high-intensity genetic sequencing methods that could produce or analyse that much data, much less make sense of it. But AI VIVO’s revolutionary methods have produced promising results in just 15 days.
We’ve compiled a list of top 31 drugs that might just give our cells a chance at beating Covid-19. Our team has good reason to believe their findings are accurate, because five of the drugs identified by AI VIVO have already been put into clinical trials by other scientific teams using different methods. That overlap makes for good science, and the other 26 drugs on the top of the list offer additional opportunities for more clinical trials to move forward. The more drugs we have in trials, the better.
But how long do we have to wait? Will it be years of tricky testing and human trials before any of us see the benefits? Our results suggest we can move much faster. The top 31 drugs identified as potential treatments are ‘ready to go’.
They are already either FDA-approved for treating other human diseases, or well on their way through Phase II and Phase III trials. In a pandemic, we simply don’t have the time and billions of pounds to waste on developing new drugs from scratch.
We have to work with what we’ve already got, quickly repurposing drugs we already know to be safe for human use. AI VIVO’s top 31 were selected from over 90,000 possible drugs, and we will be expanding the search in the coming days.
The next stage in the battle is to find a cocktail of drugs - two, three, or even more drugs - that will work together to attack Covid-19 as a team.
For example, some drugs might block the virus’s access to human cells, others might help reduce symptoms or prevent the worst outcomes, and others may simply increase the other drugs’ effectiveness. AI VIVO’s is now working hard to find the combination that does it all.
Right now, AI VIVO is busy finding partners to test their Top 31 drugs, calling for help from government, pharma and biotech companies, and research groups to push testing and trials through as quickly as possible. The international Covid-19 research effort might well be the largest scientific-medical collaboration in history, and we all depend on teams like AI VIVO who can think outside the box.
Read more
AstraZeneca begins clinical trial of blood cancer drug Calquence for Covid-19 patients
AstraZeneca targets clinical tests of new Covid-19 antibody treatments in ‘three to five months’
Healx trains AI to find Covid-19 combination therapies
