National University of Singapore (NUS) researchers used artificial intelligence to find the best combination therapies against COVID-19. Their study led them to having a combination of remdesivir, lopinavir, and ritonavir.

According to NUS News, the study is a collaboration between the NUS Department of Pharmacology, Cancer Science Institute of Singapore, Osmosis (Knowledge Diffusion), and Shanghai Jiao Tong University. The study results were published on November 10 in Bioengineering and Translational Medicine.

Artificial Intelligence helps identify best COVID-19 combination drug treatment

The NUS' N.1 Institute for Health and Institute for Digital Medicine (WisDM) Director Professor Dean Ho co-led the research team who used the ground-breaking AI technology called 'IDentif.AI' or the Optimising Infectious Disease Combination Therapy with Artificial Intelligence. BioSpectrum Asia reported that researchers tested 12 drug candidates, which are also used in others studies currently in clinical trials for treatment of COVID-19.

These drugs include remdesivir, lopinavir, favipiravir, ritonavir, hydroxychloroquine, oseltamivir, chloroquine, losartan, azithromycin, ribavirin, dexamethasone, and teicoplanin. These 12 drugs led to more than 530,000 possible combination therapies. ritonavir and lopinavir are given together for HIV treatment.

Despite the huge number of possible combination, the testing process was simplified with the use of IDentif.AI, so the selection process was already completed in just two weeks.

This showcased the effectiveness of IDentif.AI in finding optimal combination therapies for infectious diseases. "With IDentif.AI, we will always be ready to rapidly find optimal therapeutic solutions for the next outbreak," Prof. Ho noted, according to BioSpectrum Asia.

Meanwhile, the researchers is now planning to expand IDentif.AI in local therapies to develop new combinations that can be deployed and used easily, particularly for other infectious diseases.

Read also: COVID-19 Vaccine: Do You Still Have to Wear Face Masks After Getting Immunization Shots?

Using AI to find the most effective drug combinations in two weeks

Researchers used live, patient-derived SARS-CoV-2 in testing drug combinations. Traditionally, a 12-drug set will be tested with 10 different doses for each drug. This would create up to one trillion possible combinations.

However, by using IDentif.AI, the research team learned that they only need three different dosages for each drug, which cut down the possibilities to just 531,000 possible combinations. While it was still a huge number, the number of experiments needed has been significantly reduced. Thus, they were able to complete the entire study in just two weeks.

The study finds the specific doses of remdesivir, ritonavir, and lopinavir combination on top of the list, which led to nearly complete constrain of infection. Also, IDentif.AI was able to note the interaction between the three drugs, which were showcased in the experiments to increase the efficacy. This led researchers to conclude that IDentif.AI may be advantageous in realizing drug combinations that were ineffective when used alone.

"IDentif.AI is unique," Prof. Ho told NUS News. He explained that the tool allowed researchers to get the data they need by administering carefully designed experiments to create a "list of actionable and optimised regimens." This only shows how useful artificial intelligence in healthcare is.

Meanwhile, the study also named remdesivir as the best performing single drug treatment against coronavirus. In contrast, the combination of azithromycin and hydroxychloroquine, which were widely studied, was proved to be ineffective, which contradicts vast majority of earlier studies that used very high dosage in treating COVID-19. This became toxic for patients, which led to deaths of more patients as recent clinical results showed.

Related article: New CRISPR-Based Covid-19 Test Can Provide Accurate Results in Less Than 5 Minutes Using Smartphone Camera

This is owned by Tech Times

Written by CJ Robles

ⓒ 2021 All rights reserved. Do not reproduce without permission.