Different software processing affects the peak picking and metabolic pathway recognition of metabolomics data

Area of Science:

• Metabolomics
• Clinical Chemistry
• Biotechnology

Background:

• Untargeted liquid chromatography-mass spectrometry (LC-MS) is vital for identifying disease-related metabolic pathways.
• Effective data preprocessing and pathway recognition are critical steps in metabolomics studies.
• Evaluating different LC-MS methods is essential for reliable clinical sample analysis.

Purpose of the Study:

• To compare the performance of three common LC-MS data processing software (XCMS, Progenesis QI, MarkerView) for coronary heart disease (CHD) serum samples.
• To assess the impact of different chromatographic columns (BEH amide and C18) on metabolomics data quality and pathway analysis.
• To provide guidance for selecting optimal methods in clinical metabolomics research.

Main Methods:

• High-resolution mass spectrometry data were collected from 221 CHD patients.
• Serum samples were analyzed using both BEH amide and C18 chromatographic columns.
• Data preprocessing and metabolic pathway enrichment were performed using XCMS, Progenesis QI, and MarkerView, with StatTarget for correction.

Main Results:

• All three software programs exhibited varying degrees of signal drift, but StatTarget improved data quality.
• XCMS demonstrated superior performance in detecting real chromatographic peaks.
• Progenesis QI achieved the highest number of metabolite annotations.
• Complementary performance was observed between C18 and amide columns for metabolic pathway analysis.

Conclusions:

• XCMS and Progenesis QI offer distinct advantages in LC-MS data processing for metabolomics.
• Combining XCMS and Progenesis QI, along with complementary chromatographic columns, enhances the reliability of metabolic pathway identification.
• This study provides valuable insights for optimizing LC-MS-based metabolomics in large clinical cohorts.