Improved Chromatographic Alignment of GC-IMS Data Using Spectral Information

Gas chromatography-ion mobility spectrometry (GC-IMS) provides two-dimensional separation of volatile compounds, but instrumental variability can produce shifts and nonlinear distortions that hinder comparison among measurements. This study presents a chromatographic alignment workflow that uses spectral information from the drift-time dimension to support peak correspondence. The method combines injection point correction and parametric time warping along the retention-time axis with multiplicative correction of the drift-time axis. Peaks are then matched using information from both dimensions, and their retention positions are used as internal landmarks for a final cubic spline refinement without requiring standards specifically for alignment. The workflow was evaluated using 134 repeated measurements of a pooled urine sample and compared with correlation optimized warping (COW) applied directly and after different preprocessing stages. Conventional metrics based on correlation, Euclidean similarity, simplicity, and peak factor indicated comparable global performance for the proposed method and the best COW workflows. However, a correspondence-based assessment showed that high profile similarity and low positional dispersion did not necessarily imply preservation of the same peak correspondences. Direct COW recovered 45.2% of the reference correspondences, whereas recovery increased to 85.0% after injection point correction and to 89.9% after injection point correction plus parametric time warping. These results highlight the importance of combining profile-based metrics with measures of peak correspondence and show how the orthogonal drift-time dimension can provide additional constraints for chromatographic alignment in GC-IMS data.