Algorithm for comprehensive analysis of datasets from hyphenated high resolution mass spectrometric techniques using...

Area of Science:

• Analytical Chemistry
• Computational Chemistry
• Biochemistry

Background:

• Hyphenated mass spectrometry (MS) generates complex data requiring advanced algorithms for information extraction.
• Extracted ion profiles are crucial for identifying separated compounds but can be challenging due to analyte-specific signals and multiple m/z ranges per compound.
• Existing methods struggle with visualizing poorly separated compounds or those with vastly different intensities.

Purpose of the Study:

• To develop and validate a novel hierarchical agglomerative clustering algorithm for grouping extracted ion profiles in hyphenated MS data.
• To enable more comprehensive analysis and visualization of complex MS datasets, including challenging low-intensity or poorly resolved analytes.
• To improve the identification and characterization of compounds from complex mixtures.

Main Methods:

• Implementation of a hierarchical agglomerative clustering algorithm to group similar extracted ion profiles.
• Generation of a 2D "clusters plot" for in-depth MS dataset analysis and visualization.
• Validation using capillary electrophoresis time-of-flight mass spectrometry (CE-TOF-MS) with electrospray ionization.

Main Results:

• The algorithm successfully clustered ionic profiles belonging to individual compounds, even with low separation (R < 0.03).
• Analysis of a urine sample yielded 70 clusters from a total ion profile showing only 15 peaks, demonstrating enhanced analytical depth.
• The clustering approach effectively visualized poorly separated compounds with intensity differences exceeding two orders of magnitude.
• Computational time for analysis was less than 10 minutes.

Conclusions:

• The hierarchical agglomerative clustering algorithm provides a robust method for analyzing complex hyphenated MS data.
• This approach significantly improves the detection and characterization of analytes, especially in complex biological samples like urine.
• The clusters plot offers a powerful visualization tool for in-depth understanding of MS datasets.