This project combines the fields of biochemistry and data science by applying a machine learning approach to the identification of potential metallo-β-lactamase (MBL) inhibitors. MBLs are enzymes expressed by antibiotic-resistant bacteria and are becoming more clinically prevalent, leading to an increasing number of severe cases of once easily-treatable infections. In recent years, there has been remarkable effort put into identifying MBL inhibitors but, to date, there are no clinically-viable MBL inhibitors. We developed a novel, iterative approach to MBL inhibitor identification, using an initial computer model based on previous inhibition data that ranks compounds from large chemical libraries based on their likelihood of success as an MBL inhibitor. The top 2,816 compounds from the current iteration of the model were tested via high-throughput screening and 851 had a meropenem or nitrocefin AC50 value less than 10 μM, equating to a success rate of 30.2%. The model will be continuously improved, with plans to incorporate data from high-throughput screening experiments conducted on the NDM-1, VIM-2, and IMP-1 variants of MBL. This updated model will be used to screen more diverse chemical libraries and further biochemical and qHTS studies will be performed on “hits” identified by the final model. Additional filters based on physiochemical properties will also be integrated into the model. Future work will include the development of a generative model that will use data from the updated model to generate novel MBL inhibitors. This project has exemplified interdisciplinary research, showing us the intersection of chemistry and data science. It has shown that data science is not limited to traditional uses and can be applied to medically-relevant issues, as well. All three of us will seek to apply data science to all facets of our future careers, whether they be in data science, business, or medicine.
Authors: Aidan M. Sturgill, Mitchell R. Fairweather, Amy Hu
Faculty Advisors: Michael W. Crowder, Chemistry & Biochemistry; Maria L. Weese, Information Systems & Analytics; Waldyn G. Martinez, Information Systems & Analytics
Graduate Student Advisors: Zishou Cheng and Caitlyn A. Thomas, Chemistry & Biochemistry
You must be logged in to post a comment.