AI-based COVID-19 Drug Discovery
The speed and the disruptive nature of the COVID-19 pandemic have taken both public health and biomedical researchers by surprise, demanding the rapid deployment of new interventions, as well as development and testing of an effective cure and vaccine. Considering the compressed timescales, the traditional methodologies relying on iterative development, experimental testing, clinical validation, and approval of new compounds are not feasible. A more realistic strategy relies on drug repurposing, required to identify clinically approved drugs, with known toxicities and side effects, that may have a therapeutic effect in COVID-19 patients.
Network medicine approaches can be used to study host-pathogen interactions, helping illuminate infection mechanisms and identify new treatment strategies. These tools are currently being applied to analyze the perturbations to intracellular networks as COVID-19 invades the cell. This information is in turn being analyzed to identify approved drugs that could be rapidly repurposed to treat COVID-19. A partnership between Northeastern University’s Barabasi Labs, Harvard Medical School, the Network Science Institute and biotech start-up Scipher Medicine is on the search for drugs that can quickly be repurposed as Covid-19 treatments. They have thus far mapped out and characterized the “network neighborhood” of the human interactome perturbed by SARS-COV2 (COVID-19 disease module), and have explored four network-based strategies to prioritize existing drugs based on their ability to perturb this disease module. Already, a draft list of 20 top predictions has been generated.
In this invention, Northeastern University researchers have adapted the network-based toolset to COVID-19, recovering the primary pulmonary manifestations of the virus in the lung as well as observed comorbidities associated with cardiovascular diseases. Also, the research results predict that the virus can manifest itself in other tissues, such as the reproductive system, and brain regions, and could have neurological comorbidities.
Northeastern University researchers build on these findings to deploy three network-based drug repurposing strategies, relying on network proximity, diffusion, and AI-based metrics, allowing to rank all approved drugs based on their likely efficacy for COVID-19 patients, by aggregating all predictions the result arrived with 81 promising repurposing candidates. The drugs currently in clinical trials are used to validate the accuracy of these predictions, and an expression-based validation of selected candidates suggests that these drugs, with known toxicities and side effects, could be moved to clinical trials rapidly.
- Output list of ranked drugs can have a therapeutical effect for treating COVID-19
- Combination of network methods is novel, using AI-Network methods, graph theory proximity-based models, and diffusion methods
- Can find potential drugs and fast-forward clinical trials for COVID‑19 treatment
- Deisy Morselli Gysi, Albert-Laszlo Barabasi, Italo do Valle, Onur Varol, Xiao Gan, Asher Ameli, Joseph Loscalzo, Marinka Zitnik
- Covid-19 Drugs