Untargeted Chemical Profiling of Two-Dimensional Gas Chromatography Coupled with High-Resolution Mass Spectrometry...

Area of Science:

• Analytical Chemistry
• Chemometrics
• Data Science

Background:

• Advanced chemical profiling of complex samples like botrytized wines demands sophisticated analytical techniques.
• Conventional methods struggle to interpret the vast data generated by techniques such as comprehensive two-dimensional gas chromatography (GC × GC) coupled with high-resolution time-of-flight mass spectrometry (HR-TOF-MS).
• Subtle compositional variations in natural matrices are often missed due to data volume and complexity.

Purpose of the Study:

• To introduce a statistically robust framework for untargeted identification of sample-specific chemical markers in complex samples.
• To leverage a Topological Data Analysis (TDA) tool, Ball Mapper, for processing and interpreting large-scale GC × GC-HR-TOF-MS data.
• To demonstrate the framework's capability in characterizing chemical signatures reflecting regional and process-related differences in wines.

Main Methods:

• Development of a statistically robust framework integrating Topological Data Analysis (TDA) with Ball Mapper.
• Application of the framework to analyze data from comprehensive two-dimensional gas chromatography (GC × GC) and high-resolution time-of-flight mass spectrometry (HR-TOF-MS).
• Discretization, normalization, and clustering of over 470,000 mass spectral functions per sample to generate discrete mass spectral vectors.

Main Results:

• The framework successfully processed immense data volumes, identifying 2,792 extracted mass spectral distributions and 1,191 discriminative features.
• 334 compounds were assigned to known volatile organic compound classes, revealing significant chemical signatures.
• The analysis reflected regional differences, fermentation styles, grape varieties, botrytization conditions, and microbial activity, with consistent grouping patterns observed.

Conclusions:

• The proposed statistically robust framework using Ball Mapper is effective for the chemical characterization of complex natural matrices.
• This approach provides a general strategy for analyzing large datasets generated by GC × GC with HR-TOF-MS across various domains.
• The method enhances the interpretation of complex chemical data, enabling the discovery of subtle yet significant chemical markers.