Harnessing data to combat the pandemic

Data, whether they are of an IT, medical or other nature, reflect the multiple facets of the current health crisis and its associated challenges. Data can be found at every stage of the fight against the SARS-CoV-2 virus, from situation analysis up to decision-making, and including the development of treatments, the management of human and material resources, and more. On the basis of this fact, we at IVADO are supporting nine projects that are part of the race against time that began worldwide in March 2020.

Let’s go back 11 months: while the health crisis was disrupting our daily lives, we were launching an unsolicited call for proposals to combat COVID-19. Our IVADO community mobilized quickly and nine projects obtained funding and resources to search for solutions. Let’s take a look at these initiatives!

Multiple databases are being developed worldwide because of the pandemic. David Ardia (HEC Montréal) and Emanuele Guidotti (Université de Neuchâtel) immediately grasped the importance of collecting and accessing these data. In the spring of 2020, they launched COVID19 Data Hub, an international open-access platform that brings together more than 180 countries. It has already been downloaded more than 3.4 million times and won the “CovidR” competition of the 2020 European R Users Meeting (eRum). It provides “clean” ground for collecting and analyzing a large amount of reliable and diversified data (clinical data, data from public policies, etc.), in order to enhance our understanding of the crisis. But the situation also alerted us to the need for foresight!

Indeed, globalization-related phenomena, such as deforestation, could lead to other pandemics similar to those we are currently experiencing if we do not take appropriate measures. We have realized how animals can serve as hosts for all kinds of viruses that can be transmitted to humans: it is therefore essential to be able to identify these pathogen-reservoir animals quickly, and here again, data are our allies. Timothée Poisot (Université de Montréal) and Colin Carlson (Georgetown University, guest researcher at Université de Montréal) developed algorithms to predict these potential hosts and managed to develop a reliable model that can be extended to other viruses. We now had a prevention tool to better detect the emergence of new viruses.

Back to the present time: researchers in every country are trying to establish the origin of the virus within their national territory, in order to be able to study and track data on its different mutations. These studies aim to better understand how—and how fast—the virus spreads, and to estimate its pathogenicity. In the Canadian province of Québec, David Stephens (McGill University) and Luc Villandré (HEC Montréal) have so far identified 20 case clusters.

By linking these data to patient data, it is possible to establish better prognoses, to understand why certain populations present different symptoms after infection with the virus, and consequently to predict which people are most at risk. Julie Hussin (MHI [Montréal Heart Institute], Université de Montréal) uses bioinformatics tools and genomic profiling methods to identify virus subtypes and monitor the evolution of the virus in real time.

Time is of the essence… In the COVID19 battle against the clock, the capacity to test populations is another crucial issue. PCR tests require reagents and logistics that take time to be deployed. Frédéric Leblond (Polytechnique Montréal) and Dominique Trudel (CHUM [Université de Montréal teaching hospital]) are developing very fast (20 seconds) and easy-to-use pre-tests based on a technology designed to determine an individual’s viral load. These tests would make it possible to quickly assess whether or not it is necessary to continue to use a PCR test for screening.

In addition to the diagnostic challenges, research teams around the world are mobilizing to find effective treatments to stem the crisis. The “Achilles heel” of SARS-CoV-2 must be identified in order to be able to determine the molecules capable of fighting it! There are several possible strategies…

Among them, François Major (IRIC [Institute for Research in Immunology and Cancer], Université de Montréal) an RNA specialist, is working on developing of an algorithmic model capable of predicting 3D structures of the virus’s RNA. The objective is to use these 3D structures to identify small molecules likely to interact with the virus and have a therapeutic action.

For their part, Yoshua Bengio (Mila [research institute in artificial intelligence], Université de Montréal) and Mike Tyers (IRIC, Université de Montréal) are looking at viral proteins. They have developed an algorithm based on reinforcement-learning capable of representing small synthesizable molecules. In-silico affinity tests—i.e. computer-modelled affinity tests—are under way to identify candidate molecules. The potential of this work could go as far as revolutionizing the drug discovery chain!

Whatever treatment is being considered, clinical trials should be carried out before making it available to the public. Recruiting and monitoring patients during this phase is a major challenge. An innovative process has been put in place to meet these challenges, as part of the COLCORONA study conducted by Jean-Claude Tardif (MHI) and Frédéric Lesage (Polytechnique). This is a digital platform that makes it possible to speed up the evaluation of a treatment by using electronic-consent methods, videoconferencing consultations and a support chatbot for nurses. This methodology has enabled tests to be carried out quickly, leading to the breakthroughs recently announced in the media, including in Le Devoir (in French). This platform could be used for other studies, or even generalized, in order to accelerate clinical trials while reducing costs. In the near future, there are plans to link the genetic data of the virus collected by Julie Hussin’s team to those of the patients, and to integrate them into the COLCORONA study. This will make it possible to analyze the impact of the different variants of the virus on patients and the repercussions on treatment efficacy depending on the case.

The management of hospital resources is another major pandemic-related challenge that regularly makes headlines because of the constant fear of overburdening intensive care units (ICUs). To ensure better patient monitoring and facilitate the allocation of resources by medical staff, the team of Philippe Doyon-Poulin (Polytechnique) and Philippe Jouvet (Sainte-Justine hospital, Université de Montréal) have created a numerical chart to view patient health status and hospital bed capacity. This digital tool is currently being implemented in the pediatric ICU of Sainte-Justine hospital and the ICU the Jewish General Hospital.

Through the initial outcomes of these nine complementary projects, we can see how data are involved at each stage of the crisis-resolution process, and the importance of promoting their accessibility to facilitate research work. At IVADO, we are committed to supporting the synergy between the different disciplines, in order to contribute to improving the situation and to new discoveries.

The IVADO team