Beyond Chloroquine: AI helps identify possible treatments for COVID-19

Yiannis Kiachopoulos
published on March 22, 2020

With trials for Chloroquine well under-way, at Causaly we’ve taken a deeper look into the world’s biomedical literature, uncovering additional compounds with potential to be repurposed as COVID-19 treatments.

We looked for approved sustances known to inhibit targets which increase members of the Coronaviridae family, that also have a documented link to the ACE2 protein (like Chloroquine).

In total, we identified 90,000 connections between approved drugs, potential drug targets and Coronaviridae viruses.

covid19-all-rels-with-exclusions-flat-new

42 substances were identified warranting further research. If you or someone you know is actively researching COVID-19, we’re offering free open access to this data, and Causaly AI for any non-commercial research. Contact us to request access.

Here's how Causaly AI is currently being used to accelerate COVID-19 research:

Mapping existing treatment options

Causaly is being used to rapidly identify all the treatment options for similar diseases. By machine reading existing literature, you have instant access to all previously published evidence for the betacoronavirus genus, and can easily identify those treatment options with the most supporting evidence.

This yields 131 substances that can be further investigated. Among the most prominent are Type I interferon, chloroquine and the antiviral drugs lopinavir and ribavirin, which have been characterized as potent inhibitors of the SARS-CoV and MERS-CoV spread.

Image-1_new
Image 1. Potential treatments for the betacoronavirus genus.

Understanding the disease mechanism

In addition to treatment options, Causaly AI allows researchers to find biomarker genes and potential molecular targets of a disease. In the case of the betacoronavirus genus, a total of 89 results are found.

At the top of the list, Transmembrane Serine Protease 2 (TMPRSS2) gene expression is shown to induce SARS spread.

Taking a closer look at the underlying evidence, it has been reported that the TMPRSS2 might promote the SARS-CoV spread by activating the spike protein SARS S. Notably, it has also found that the COVID-19 also uses the cellular protease TMPRSS2 for entry into target cells.

Image-2.v02

Image 2. Genes related to the betacoronavirus genus.

Identifying potential drug candidates

Once a potential molecular pathway has been identified, researchers can then identify potential pharmaceuticals.

Continuing with the same example, searching for molecules that either suppress TMPRSS2 gene expression or inhibit TMPRSS2 yields 12 drug candidates which may warrant more investigation by researchers as potential treatments.

Image-3

Image 3. TMPRSS2 protein and related pharmacologic substances.

Results of this search include camostat, the serine protease inhibitor which has already been suggested by Hoffmann et al. as a potential drug candidate for COVID-19.

Research organisations are able to inspect similar relationships for all the other genes related to the CoV molecular signaling, uncovering relationships that would not be obvious by traditional literature review search, rapidly accelerating research.

Causaly’s COVID-19 dataset has been provided to Bill and Melinda Gates' Global Health Drug Discovery Institute (GHDDI).

Request access for your organisation

If you would like access for your organisation, and are currently researching COVID-19, we’re offering free access for all non-commercial research. Contact us to request access.

How Causaly AI is transforming translational research: Interview with Imad Yassin
Application

How Causaly AI is transforming translational research: Interview with Imad Yassin

An interview with Imad Yassin on how AI is transforming translational research and why Causaly is one of the biggest game-changers in the field he's seen to date.

COVID-19: Causaly AI releases Top 250 compounds for further research following CORD-19 Network Analysis

COVID-19: Causaly AI releases Top 250 compounds for further research following CORD-19 Network Analysis

To unlock all published scientific knowledge on COVID-19, Causaly AI has performed a network analysis of the 40,000 COVID-19 full text...

Covid-19: Causaly partners with UCL to accelerate coronavirus research
News

Covid-19: Causaly partners with UCL to accelerate coronavirus research

At Causaly we're excited to announce a new partnership with University College London (UCL) to help accelerate research into the...

Sign up for Causaly newsletter