Comprehensive analysis of chromatographic data by using PARAFAC2 and principal components analysis

Area of Science:

• Analytical Chemistry
• Chemometrics
• Chromatography

Background:

• Traditional GC-MS analysis relies on peak integration and PCA, which struggle with baseline drifts and poorly defined integration boundaries.
• Instrumental artifacts and co-eluted peaks complicate accurate data interpretation in GC-MS.

Purpose of the Study:

• To propose an improved methodology for chromatographic data analysis using PARAFAC2 and PCA.
• To overcome limitations of conventional GC-MS data processing techniques.
• To enhance analyte detection and deepen the understanding of complex datasets.

Main Methods:

• Modeling raw GC-MS datasets using the PARALLEL FACTOR 2 (PARAFAC2) algorithm in specific GC profile regions.
• Utilizing the resolved chromatographic profiles from PARAFAC2 to build a subsequent Principal Component Analysis (PCA) model.
• Applying the combined PARAFAC2+PCA methodology to analyze the aroma profile of 36 ripening apples.

Main Results:

• The PARAFAC2+PCA method effectively overcomes issues related to instrumental artifacts like baseline drifts.
• Improved detection of previously unidentified analytes was achieved.
• A more comprehensive understanding of the studied GC-MS dataset was obtained.

Conclusions:

• The proposed PARAFAC2+PCA methodology offers significant advantages over traditional GC-MS analysis.
• This approach saves researcher time and effort.
• The findings are applicable to other hyphenated chromatographic datasets, demonstrating broad utility.