Fighting a virus with data
Like many of us earlier this year, Dr. Ahmed Abdeen Hamed, a Data Science Professor at Norwich University, listened to Dr. Anthony Fauci explain to reporters at a news conference, that a vaccine for the novel Coronavirus that leads to the COVID-19 disease, might be months away.
Dr. Hamed, Norwich University’s first specialist in data science has always had an interest in solving real-world problems and, at this moment, COVID-19 ranks at the top of the world’s problems. He thought of the unfortunate individuals that were already suffering from COVID-19 and, immediately, the idea of drug repurposing came to mind.
How could healthcare providers repurpose drugs that are now utilized to fight known viruses into a possible treatment for those suffering from this new disease?
Dr. Hamed then turned to PubMed, the search engine which combs through the MEDLINE database of references and abstracts on life sciences and biomedical topics.
There are thousands of articles in the database describing the SARS-CoV-2 treatment, which he thought there might be benefits for repurposing for COVID19. The challenge that Dr. Hamed faced was identifying and validating such benefits, if they exist. Dr. Hamed turned to COVID19 clinical trials to validate the drug repurposing evidence and produced a list of drugs from these trials that could be recommended for repurposing.
Enter the algorithm
How does a data scientist go about entering the fight against COVID-19? An algorithm was his weapon of choice.
As a computer scientist specializing in Data Science, Dr. Hamed designed an algorithm that analyzed the results of his search through the PubMed database. Utilizing natural language processing techniques, drugs were identified, and a map was created allowing him to dive deeper into the results with network science tools. Dr. Hamed’s CovidX algorithm was applied to the network of drugs, and a ranking was given to those drugs that were currently being investigated, validating the mechanism of the PubMed results.
Once the literature was analyzed, the algorithm identified 30 drug candidates for possible repurposing. The top 5 candidates according to the algorithm were hydroxychloroquine, azithromycin, chloroquine, ritonavir and losartan.
Making sense of it all
Dr. Hamed sought the assistance of Dr. Lyndsey Gates, a Norwich University Assistant Professor of Nursing to validate the results utilizing her medical expertise. Dr. Gates stepped in to help interpret the findings and “put the puzzle pieces together”. They both acknowledged that the ranking is not absolute as the pandemic is developing rapidly but the hope is that the higher-ranking drugs from this report can form a sort of adjuvant therapy for treating those suffering from COVID-19.
Both Dr. Gates and Dr. Hamed recorded their findings with the Journal of Medical Internet Research with an article titled: The Anatomy of the SARS-CoV-2 Biomedical Literature: Introducing the CovidX Network Algorithm for Drug Repurposing Recommendation [“https://www.jmir.org/2020/8/e21169/”]
Dr. Hamed’s research is partly supported by Norwich University Applied Research Institute (NUARI).
NUARI is a 501(c)(3) non-profit that serves the national public interest through the interdisciplinary study of critical national security issues. We are partially funded by the Department of Homeland Security and the Department of Defense and federally chartered under the sponsorship of Sen. Patrick Leahy. We are co-located with Norwich University in Northfield, VT, and share their ideals of academic excellence, innovation, and service to the country.