Selective paired ion contrast analysis: a novel algorithm for analyzing postprocessed LC-MS metabolomics data...

Area of Science:

• * Metabolomics
• * Bioinformatics
• * Biomarker Discovery

Background:

• * Analyzing biological data from human subjects is challenging due to high variability in uncontrolled factors.
• * Metabolomics, the study of small molecules, offers quantitative insights but is susceptible to noise, hindering biomarker discovery.
• * Confounding factors in biofluid samples like urine can obscure relevant biological information.

Purpose of the Study:

• * To develop a novel algorithm, Selective Paired Ion Contrast (SPICA), for extracting biologically relevant information from noisy metabolomics data.
• * To address limitations of single-ion analysis in metabolomics by introducing an ion-pair based statistical approach.
• * To enhance biomarker discovery in complex biological systems by improving data analysis rigor.

Main Methods:

• * SPICA redefines the unit of statistical analysis from single ions to all possible ion-pair combinations within a dataset.
• * Statistical comparisons are performed by analyzing differences between ion pairs, enabling normalization when single metabolites are insufficient.
• * The algorithm was applied to human urine datasets from total body irradiation (TBI) and colorectal cancer (CRC) relapse studies.

Main Results:

• * In the TBI study, SPICA identified 3530 significant ion-pairs, revealing radiation-specific metabolite-pair biomarkers linked to perturbed metabolic pathways.
• * In the CRC study, 6461 significant ion-pairs were identified, with several mapping to folic acid biosynthesis, a key pathway in colorectal cancer.
• * Support vector machines (SVMs) built using SPICA features significantly outperformed those based on classical single-ion features.

Conclusions:

• * SPICA offers a statistically rigorous method for analyzing noisy metabolomics data, outperforming conventional single-ion approaches.
• * The ion-pair analysis facilitates the discovery of novel metabolite-pair biomarkers and the elucidation of affected metabolic pathways.
• * SPICA demonstrates significant potential for advancing biomarker discovery and understanding disease mechanisms in complex biological samples.