ABSTRACT
Chromatographic methods are probably the most appreciated analytical techniques and well suited for the analysis of complex mixtures. The peak shifts that are observed in a collection of chromatograms are a major obstacle in their comparative analysis using different multivariate chemometric methods. The presence of peak shifts in signals introduces additional variability and thus has a substantial impact on the construction of different chemometric models and their final interpretation. Therefore, signal alignment is a necessary step. Many alignment techniques have been proposed to compensate for the peak shifts in chromatographic signals. A target signal plays the role of a template for the alignment, and all chromatograms are aligned with respect to the corresponding features that are found in the target signal. Correlation optimized warping (COW) is one of the oldest and most popular alignment approaches. In general, the available alignment algorithms that handle two-dimensional chromatographic signals belong to two categories: peak-based and raw data-based technique.
