Oxidative Stress Biomarkers for Rheumatoid Arthritis via Machine Learning

Figshare hosts a research dataset identifying potential oxidative stress-related biomarkers and therapeutic drugs for rheumatoid arthritis. The study analyzed transcriptome data from the GEO database, identifying 281 differentially expressed oxidative stress-related genes and three core biomarkers using machine learning algorithms. Author Zhigang Shi published the findings in March 2026.

Use Cases

• Validate the identified core biomarkers (CXCL10, EDNRB, MMP13) using the provided gene expression data from the GEO database.
• Analyze the 281 differentially expressed oxidative stress-related genes (DEOSGs) for enrichment in MAPK, AMPK, TNF, and Toll-like receptor signaling pathways.
• Investigate the predicted binding affinity of the candidate drug narciclasine with target proteins using the molecular docking results.

Strengths

• Identifies 281 differentially expressed oxidative stress-related genes (DEOSGs) from transcriptome analysis.
• Integrates results from three machine learning algorithms to determine three core biomarker genes.
• Includes validation from in vitro cell model experiments confirming bioinformatics findings.

Limitations

• Dataset is very small at 38.4 KB, indicating limited raw data or summary-level results only.
• Specific sample sizes, column structures, and raw transcriptome data rows are not provided.
• Findings are preliminary and based on computational screening, requiring further experimental validation.

Provenance

Source

Transcriptome data sourced from the GEO database.

Collection Method

Integrative bioinformatics analysis combining differential expression, weighted gene co-expression network analysis, and three machine learning algorithms.

Freshness

Last updated March 2026.