Developing 2D mzCompare for single comprehensive two-dimensional chromatography time-of-flight mass spectrometry...

Accurate identification of all detectable analyte components in a single comprehensive two-dimensional (2D) gas chromatography time-of-flight mass spectrometry (GC × GC-TOFMS) chromatogram is a fundamental interest in the field. While commercial software tools intended for this purpose are available, the performance of these tools to generate an accurate peak table has generally not been validated. To address this, we developed a new algorithmic software approach called 2D mzCompare to generate accurate peak tables for GC × GC-TOFMS. Extending from our original method for one-dimensional GC-MS data, the 2D mzCompare algorithm discovers selective mass channels (m/z) for each analyte to resolve overlapping peaks and improve analyte identification, by leveraging the similarity in retention time and peak shape across m/z of the same analyte. The 2D mzCompare algorithm calculates the peak shape similarity between m/z at every modulation, followed by clustering and focusing steps, to generate a final peak table. To evaluate this software, we simulated realistic GC × GC-TOFMS data in the context of the statistical overlap theory (SOT), so the exact number and identities of analytes are known a priori. Utilizing an in-house mass spectrum library of similar compounds, GC × GC-TOFMS chromatograms were simulated with varying degrees of 2D chromatographic saturation (α_2D). At low saturation factors (α_2D = 0.01, 0.03, and 0.1), over 95% of the simulated components are found to be mathematically resolved singlets (pure analyte components) by 2D mzCompare. Meanwhile, approximately 62% were found at α_2D = 1, exceeding predictions made by SOT. Application of 2D mzCompare computationally reduces the ²D peak widths, resulting in an ∼12-fold reduction in α_2D. The results of this research are simultaneously two-fold. First, we provide a new algorithmic approach, 2D mzCompare, to resolve overlapped analytes in GC × GC-TOFMS data, and second, we validate the accuracy of the software performance using SOT.