Pediatric Sepsis Biomarkers Identified via Machine Learning

Gene expression data and identified biomarkers for pediatric sepsis, derived from six integrated RNA-seq datasets covering 497 patients and 116 healthy controls. The dataset was created by Weidong Ye and last updated in April 2026. It includes 237 high-confidence biomarkers, with RORA and GPR183 highlighted across multiple models.

Use Cases

• Validate biomarker candidates like RORA and GPR183 for sepsis diagnosis using gene expression profiles.
• Analyze correlations between biomarker expression levels and immune cell infiltration data, such as CD8+ T cell counts.
• Benchmark the performance of the 14 referenced machine learning models for biomarker identification on this cohort.
• Investigate the functional roles of identified biomarkers in biological processes like transcription, translation, and cellular senescence.

Strengths

• Integrated data from six distinct RNA-seq studies, providing a consolidated cohort.
• Includes 497 pediatric sepsis patient samples and 116 healthy control samples for comparison.
• Identifies 237 high-confidence biomarkers validated through multiple analytical methods.

Limitations

• The dataset is small in file size (292.6 KB), indicating limited raw data or summary-level results only.
• Specific gene expression values, sample-level metadata, and model implementation details are not provided in the description.
• Geographic and demographic details of the patient cohort are unknown, limiting generalizability assessment.

Provenance

Source

figshare, authored by Weidong Ye.

Collection Method

Integration and analysis of six bulk RNA-seq datasets from prior studies using weighted gene co-expression network analysis and machine learning.

Time Range

null

Freshness

Last updated in April 2026.

Geography

null