Mass Spectrometry and Machine Learning Applied to Abiotic and Biotic Degradation of Dental Composites

The first part of this thesis involves the in vitro study of chemical degradation products leached from synthetically prepared dental resin composites into artificial saliva (AS) and AS with esterase enzyme aging solutions. The objective of this study was to understand the degradation mechanism of dental composites through elution studies to allow the design of composites with greater chemical stability. Abiotic aging solutions were analyzed using LC-MS/MS targeting the compounds commonly found in dental composites. The highlight of this thesis is the development of a machine learning model using logistic regression to return identification probability for each compound matched to the spectral database. This machine learning model assigns single numeric probability value to each identified compound, making the results easier to interpret than having to take multiple software output values into consideration. The second part of this thesis studied the differential metabolomics of S. mutans biofilm grown on dental composites when the growth media was supplemented with glucose compared to sucrose using LC-MS. This study aimed to identify metabolites of S. mutans and understand how sugar source affects its metabolite profile. The effect of incubation time (one week, two week and one month) and biofilm sampling method (scrapping and laser ablation sample transfer) on the metabolite profile of S. mutans were also investigated.