Table 1_Integrated metabolomics and machine learning identify predictive biomarkers via SH

Wenxiu Tian published a metabolomics dataset on figshare in 2026 investigating predictive biomarkers for sintilimab-induced rash in lung cancer patients. The data includes clinical and metabolomic profiles from 55 patients, with 32 developing a rash and 23 matched controls. Untargeted metabolomic analysis of plasma was performed using UHPLC–MS/MS, identifying 92 differentially expressed metabolites.

Use Cases

• Identify predictive biomarkers for drug-induced rash based on differential metabolite analysis.
• Train machine learning models for adverse event risk stratification based on clinical and metabolomic features.
• Investigate biological pathways like oxytocin signaling or FcεRI signaling implicated in immune-related adverse events.
• Validate potential therapeutic targets for managing sintilimab-induced rash based on highlighted metabolites.

Strengths

• Includes data from 55 lung cancer patients, with 32 cases and 23 matched controls.
• Untargeted metabolomic analysis performed using UHPLC–MS/MS, a high-resolution technique.
• Identified 92 differentially expressed metabolites and 25 with high predictive value (AUC ≥0.800).
• Cross-comparison via SHAP analysis highlighted five overlapping metabolites of interest.

Limitations

• Dataset is very small at 34.0 KB, indicating limited scope and sample size.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale modeling.

Provenance

Source

Wenxiu Tian via figshare

Collection Method

Blood samples collected from lung cancer patients before sintilimab infusion and at rash onset, analyzed via UHPLC–MS/MS.

Freshness

Last updated 2026-06-03 05:48:24; freshness should be verified.