Nucleic acid interaction plays a crucial part in many common biological mechanisms, ranging from protein synthesis to cell replication. Hydrogen-bonding between complementary nucleic acid base pairs, also known as hybridization, is one of the most prominent interactions; the unique chemistry of these nucleic acid bases allow for precise yet reversible pairing. By characterizing the thermodynamic and kinetic mechanisms behind nucleic acid hybridization, it is possible to predict and quantitate the progress of nucleic acid hybridization at chemical equilibrium, or even before reaching equilibrium. In this thesis, I will explain my understanding, characterization, and prediction of nucleic acid hybridization systems, as well as how this knowledge can improve the design of DNA-based molecular diagnostics. The first two projects revolve around the double-stranded probe known as the “toehold” probe and how it can be modified to span a dynamic range of multiple orders of magnitude without the use of enzymes; I also explain my attempt to construct a probe system that could theoretically discriminate targets with different numbers of mismatches. The last two projects explore the previously uncharted territory of DNA hybridization kinetics, building predictive machine learning algorithms based of experimental hybridization kinetic data, and scaling this towards predicting kinetics in highly multiplexed settings.