Drug Repositioning in Search of COVID-19 Treatments
Dr. Guowei Wei is an MSU foundation professor in the Department of Mathematics with courtesy appointments in the Departments of Biochemistry & Molecular Biology and Electrical & Computer Engineering. Professor Wei's work focuses on reducing the geometric complexity of macromolecules so that computers can simulate how new drugs might interact with them in the human body.
In January of this year, when Dr. Wei and his team realized the potential for COVID-19 to become a pandemic, they turned their efforts towards drug discovery for COVID-19, specifically focusing on antibody therapies and small-molecule drugs. Wei's group is simultaneously investigating the mutation impact of drug discovery for the treatment of COVID-19 outbreak.
Initially, the team's research focused on drug repositioning. Since SARS-CoV-2 shares many important traits with SARS-CoV (nearly identical main protease inhibitor binding sites), all potential anti-SARS-CoV chemotherapies are also potential COVID-19 drugs. Since the development of a new drug usually takes more than ten years, instead of trying to design new drugs, Wei's team investigated whether existing drugs could be used for COVID-19 treatment. After months of working around the clock, Dr. Wei and his team have seen some success. Their work is published in the Journal of Physical Chemistry Letters.
Wei’s work in drug discovery leverages insights from artificial intelligence, advanced mathematics, and genome sequences. In terms of artificial intelligence, machine learning and deep learning are used to get a better understanding of the features and characteristics of over tens of thousands of protein-ligand complexes. Advanced mathematics (algebraic topology, differential geometry, and graph theory) is used to simplify the structural complexity of biomolecules, which are otherwise too difficult to model due to their existence in excessively high-dimensional spaces. A Mutation Tracker is being built to track the mutations from more than 35,000 SARS-CoV-2 genome isolates and understand the impact of SARS-CoV-2 mutations on diagnostics, vaccines, antibodies and drugs. Pairing advanced machine learning models and artificial intelligence with mathematical approaches and genotyping creates a formidable tool in this area of drug discovery and drug repositioning.
While the group has seen success, their research has not come without roadblocks, the biggest challenge being a lack of available data. Large volumes of data are required to utilize machine learning and deep learning to its greatest potential. In the case of SARS-CoV-2, there isn’t much data available since the virus is still relatively new and largely understudied. Nevertheless, making use of data currently available, HPCC has been incredibly useful for speeding up this work. Dr. Wei says, “given the computational power required for machine learning and deep learning, our research couldn’t have moved forward as quickly and efficiently as it has without this resource.”
Moving forward, Dr. Wei is confident that his work will play a role in controlling COVID-19 and getting life back to normal.