KPIC2: An Effective Framework for Mass Spectrometry-Based Metabolomics Using Pure Ion Chromatograms

Area of Science:

• Metabolomics
• Analytical Chemistry
• Bioinformatics

Background:

• Accurate metabolite quantitation from LC-MS data is vital for biomarker discovery but challenging due to biological complexity.
• Existing toolboxes often lack a comprehensive workflow for LC-MS data processing based on pure ion chromatograms (PICs).

Purpose of the Study:

• To develop KPIC2, an integrated R framework for robust metabolomics data analysis using PICs.
• To enhance the accurate detection, alignment, grouping, and quantitation of ions in LC-MS datasets.

Main Methods:

• KPIC2 utilizes pure ion chromatograms (PICs) for accurate ion extraction and alignment across samples.
• The framework incorporates functionalities for isotope/adduct grouping, missing peak imputation, and multivariate pattern recognition.
• Performance was evaluated against XCMS and MZmine2 using MM48, metabolomics quantitation, and soybean seed datasets.

Main Results:

• KPIC2 demonstrated superior extraction of true ions with reduced feature detection compared to other methods.
• The framework exhibited strong quantitative performance on a dedicated metabolomics quantitation dataset.
• Satisfactory classification results were achieved on a soybean seed dataset using kernel-based OPLS-DA and random forest.

Conclusions:

• KPIC2 provides a powerful and integrated solution for LC-MS-based metabolomics data analysis.
• The open-source R package offers improved accuracy and efficiency for metabolite quantitation and biomarker identification.
• KPIC2 facilitates advanced statistical analysis and pattern recognition in complex biological samples.