Retroactive Curiosity Within Your Data: Asking Your Multidimensional Gas Chromatographic Samples Different Questions

The use of comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry (GCxGC-TOFMS) is becoming more commonplace in a wide range of applications and industries. The increase in information about a given sample vastly increases with the higher peak capacity and detectability of analytes afforded by the technique. Given the full profile of analytes within a complex matrix, it is often possible to answer the research question more effectively and in more detail than with traditional one-dimensional GC-MS. However, the wealth of data available from this type of nontargeted analysis may make it possible to retroactively mine for the answers to additional questions within the data. This presentation will focus on applying data processing strategies to probe food and forensic samples for additional understanding than their original research intent. The application of different statistical approaches (i.e. supervised vs. unsupervised), grouping strategies (timeline data, sample types, experimental/control), and feature comparison metrics (F-Ratio, Coefficient of Variation, Log2Fold Change) will be discussed for real data sets. Samples initially monitored in fully nontargeted mode with the goal of capturing all possible analytes can later be compared in numerous ways to uncover key features of different categories within the dataset.