Exploratory analysis of untargeted metabolomics datasets from U-BIOPRED project

Abstract

Introduction Asthma is a heterogeneous disease with a complex pathophysiology. There is an urge to improve the understanding of asthma mechanisms and disease management at the molecular level. This study aims to identify the most altered features associated with asthma, followed by discovering asthma-related urinary metabotypes and differentiators. Methods From the cross-sectional U-BIOPRED project, adult baseline urine samples were collected from healthy participants (n=95), patients with mild-to-moderate asthma (n=84), non-smoking severe asthmatics (n=293), and smoking/ex-smoking severe asthmatics (n=101). Untargeted metabolomics data were measured using high-resolution mass spectrometry. Univariate analysis and consensus clustering methods were used for statistical analysis. Results From two complementary HILIC methods, 247 compounds were tentatively identified, composed of 205 metabolites from ZIC-HILIC positive ionization mode and 42 from negative ionization mode. Four subgroups with different metabolic patterns were classified. Endogenous steroid metabolites, caffeine metabolites, and carnitine species differentiated asthmatic subgroups from healthy controls. Conclusions A total of 233 unique urinary metabolites were identified from two complementary ZIC-HILIC ionization modes. Although the unsupervised clustering algorithm yielded subgroups with different metabotypes, the differentiated metabolites are not specific enough to conclude any phenotypes or molecular descriptors.