Effectiveness of Global, Low-Degree Polynomial Transformations for GCxGC Data Alignment

Area of Science:

• Analytical Chemistry
• Chromatography

Background:

• Retention times in comprehensive two-dimensional gas chromatography (GCxGC) can vary due to column aging and system differences.
• Accurate analysis of GCxGC chromatograms requires aligning retention times of features between runs.

Purpose of the Study:

• To assess the experimental performance of global, low-degree polynomial alignment methods for GCxGC data.
• To compare global alignment methods with a robust local alignment algorithm for GCxGC data.

Main Methods:

• Evaluation of affine, second-degree, and third-degree polynomial transformations for global alignment.
• Comparison with a local alignment algorithm using root-mean-square (RMS) residual differences for matched peaks.

Main Results:

• Global polynomial transformations outperformed the local algorithm with a sufficient number of alignment points, achieving over 95% improvement in misalignment.
• The local method showed lower error rates with small alignment point sets, despite higher computational cost.
• Neither global nor local methods performed well on pairs with slight initial misalignment, sometimes worsening the alignment.

Conclusions:

• Global, low-degree polynomial transformations are effective for aligning GCxGC chromatograms, especially with ample alignment data.
• The choice of alignment method (global vs. local) depends on the number of available alignment points and the degree of initial misalignment.
• In cases of minimal misalignment, refraining from alignment may be preferable to avoid degrading data quality.