| Two-dimensional gas chromatography (GC x GC) coupled to time of flight mass spectrometry (TOFMS) is a powerful separation tool for analyzing complex samples. This instrumentation produces complex data that requires sophisticated approaches to data analysis in order to attain useful information from the raw data. This dissertation provides a discussion of various data analysis tools. Several applications are also highlighted to demonstrate both the rationale behind data analysis tool selection and the important information these techniques can provide. Two experiments utilizing yeast, Saccharomyces cerevisiae , are described in which metabolite levels are monitored. Differences between a mutant (snf1Delta), known to be involved at the transcriptional level in the adaptation process of yeast to changes in glucose levels, and a wild type (wt) strain were monitored over time as environmental glucose was depleted. A second experiment was performed to monitor metabolite differences between 5 different strains (snf1Delta, cat8Delta, adr1Delta, adr1Deltacat8Delta, and wt) that are all important in altering transcription to adapt to changing glucose levels. The same data analysis tools can be applied to other sample types as is demonstrated with a food industry sample, cacao beans. A solid phase micro-extraction (SPME) method is developed to sample volatile compounds above a cacao bean sample. A major issue for the chocolate making industry is the detection and determination of bean quality prior to investing time and money into production. This is an issue of both quality control and food safety. Utilizing these tools it was possible to identify specific compounds that differed between good quality beans and those that had been damaged by moisture. Visible mold is an obvious sign of moisture damage once it has occurred, so good quality beans were monitored over time to track changes in specific analytes as moisture damage was incurred. This allowed for the detection of moisture damage, through changes in specific compounds, prior to visual observation of mold. Finally, a technique utilizing 13C labeled analytes is developed and demonstrated that provides quantitative information for an analyte of interest, in the presence of interfering compounds, with only a single injection. |
