[Gas chromatography-mass spectrometry data analysis algorithm based on sparse model and its application in resolving...

Area of Science:

• Analytical Chemistry
• Spectroscopy

Context:

• Gas Chromatography-Mass Spectrometry (GC-MS) is a powerful analytical technique.
• Analyzing complex mixtures often results in overlapped chromatographic peaks, complicating data interpretation.
• Accurate deconvolution of mass spectra is crucial for identifying individual components.

Purpose:

• To propose a novel algorithm for resolving overlapped peaks in GC-MS data.
• To improve the accuracy and efficiency of mass spectral deconvolution.
• To provide a robust solution for analyzing complex GC-MS datasets.

Summary:

• A step-by-step mass spectral retrieval strategy is employed, starting with efficient indexing for rough selection.
• Subsequent filtering rules like "strong peak out with high probability" and "extrusion" refine the spectral library.
• A sparse model-based regression algorithm is utilized to solve for chromatographic response values, avoiding over-fitting and extracting spectral structure.

Impact:

• The proposed algorithm demonstrates higher accuracy in resolving overlapped GC-MS peaks compared to traditional methods.
• It effectively reduces the size of the residual reference spectrum set.
• The sparse model shows satisfactory performance in analyzing severely overlapped peaks, offering a significant advancement for GC-MS data analysis.