Biomarker identification and pathway analysis of rheumatoid arthritis based on metabolomics in combination with ingenuity pathway analysis

Rheumatoid arthritis (RA) is a common autoimmune and inflammatory disease worldwide, but understanding its pathogenesis is still limited. In this study, plasma untargeted metabolomics of a discovery cohort and targeted analysis of a verification cohort were performed by gas chromatograph mass spectrometry (GC/MS). Univariate and multivariate statistical analysis were utilized to reveal differential metabolites, followed by the construction of biomarker panel through random forest (RF) algorithm. The pathways involved in RA were enriched by differential metabolites using Ingenuity Pathway Analysis (IPA) suite. Untargeted metabolomics revealed eighteen significantly altered metabolites in RA. Among these metabolites, a three-metabolite marker panel consisting of L-cysteine, citric acid and L-glutamine was constructed, using random forest algorithm that could predict RA with high accuracy, sensitivity and specificity based on a multivariate exploratory receiver operator characteristic (ROC) curve analysis. The panel was further validated by support vector machine (SVM) and partial least squares discriminant analysis (PLS-DA) algorithms, and also verified with targeted metabolomics using a verification cohort. Additionally, the dysregulated taurine biosynthesis pathway in RA was revealed by an integrated analysis of metabolomics and transcriptomics. Our findings in this study not only provided a mechanism underlying RA pathogenesis, but also offered alternative therapeutic targets for RA.

Automated summary

This study revealed the pathogenesis of rheumatoid arthritis through metabolomics, constructed a biomarker panel for accurate prediction, and identified dysregulated metabolic pathways associated with RA.