Multi-Omics and Machine Learning Data for Insomnia-Aggravated Sepsis-Induced Lung Injury

A Mendelian randomization analysis identified insomnia as a causal determinant for sepsis susceptibility. This 5.8 MB dataset, authored by Jinquan Zhang and last updated in June 2026, integrates multi-omics approaches and machine learning to identify hub genes linking insomnia to acute lung injury. Machine learning techniques, including Random Forest, SVM, and KNN, were used to refine a signature of 1,294 co-dysregulated genes down to three robust candidates.

Use Cases

• Validate causal links between sleep disorders and sepsis using Mendelian randomization results.
• Train machine learning models for risk prediction based on multi-omics biomarker signatures.
• Perform functional enrichment analysis on pathways like JAK/STAT3 signaling and chemokine signaling.
• Investigate macrophage-specific gene expression patterns using single-cell RNA sequencing data.
• Prioritize diagnostic and prognostic biomarkers, such as PTPN6, for experimental validation.

Strengths

• Includes experimental validation from in vivo and in vitro models to support computational findings.
• Employs multiple machine learning models (Random Forest, SVM, KNN) with SHAP for interpretability.
• Integrates several analysis techniques: Mendelian randomization, WGCNA, scRNA-seq, and functional enrichment.

Limitations

• The primary data file is a 5.8 MB PDF, which likely contains summarized results rather than raw, reusable data tables.
• Column-level documentation and sample data are unavailable, making direct reuse for ML challenging.
• Row count and specific data formats for the underlying omics data are unknown.

Provenance

Source

Jinquan Zhang via figshare

Collection Method

Integrative multi-omics analysis and machine learning applied to preclinical models and genetic data.

Freshness

Last updated 2026-06-01 04:17:32