Plasma Metabolomic and Lipidomic Profiles for Rheumatoid Arthritis Treatment Response

A study by Jiani Chen analyzed plasma samples from 106 rheumatoid arthritis patients and 10 healthy controls using liquid chromatography–mass spectrometry. The analysis tentatively annotated 2,279 metabolites and 2,987 lipids, identifying 12 core features associated with treatment non-response. Machine learning models using these features achieved area under the curve values exceeding 0.90 in internal cross-validation.

Use Cases

• Train classifiers to predict rheumatoid arthritis treatment response based on identified lipid and metabolite features.
• Validate potential biomarkers for disease activity using the annotated metabolomic and lipidomic profiles.
• Conduct feature selection studies on high-dimensional omics data using the LASSO-regression-refined core features.

Strengths

• Data is derived from 106 patient samples and 10 control samples, providing a clinical cohort.
• Analysis identified 12 core biomarker candidates, including specific lipids and metabolites like Cer(d18:1/16:0) and Mimosine.
• Machine learning models demonstrated preliminary high discriminatory performance with AUC values exceeding 0.90.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset size is 777.2 KB, indicating a very limited scope of data.

Provenance

Source

Jiani Chen via figshare

Collection Method

Liquid chromatography–mass spectrometry (LC-MS)-based untargeted metabolomics and lipidomics analysis of plasma samples.

Freshness

Last updated 2026-06-04 05:18:34; freshness should be verified.