Illuminating the Interface of Blocc Chemistry and Data Science: An Introduction to K-Nearest Neighbor Analysis and K-Medoids Clustering

The discovery of new chemical functions is being accelerated by blocc chemistry for iterative carbon-carbon bond formation coupled to artificial intelligence (AI) and machine learning (ML) techniques; yet, these technologies are largely absent from introductory undergraduate courses. This lab emphasizes the utility of data science tools through the context of discovering chemical function. We propose that teaching blocc chemistry provides student scientists the ability to explore functional data from frontier research generated for the purpose of this labeven with minimal knowledge of organic chemistry. Here, we report the first laboratory experiment in a sequence designed to illuminate the AI-Chemistry interface. This first experiment teaches students the fundamental principles of K-Nearest Neighbor (KNN) analysis and K-Medoids clustering to predict unknown chemical functions through the lens of blocc chemistry.