Data Sheet 1_Serum proteomic profiling of sepsis patients reveals a protein-based diagnost

85 differentially expressed proteins and 128 differential metabolites characterize the host response in sepsis caused by carbapenem-resistant Klebsiella pneumoniae. This multi-omics dataset includes serum profiles from sepsis patients, culture-negative controls, and those infected with carbapenem-susceptible or -resistant strains. Machine learning was used to construct a four-protein diagnostic panel validated in an independent cohort, with data published by Juan He in 2026.

Use Cases

• Train diagnostic classifiers based on the four-protein biomarker panel (IGHV1-8, ITGA2, PKP1, IGFBP6) described in the study.
• Analyze dysregulated host pathways like cysteine/methionine metabolism and folate-mediated one-carbon pool using integrated proteomic and metabolomic data.
• Validate proteomic signatures linked to carbapenem resistance and disease severity using the described independent cohort validation data.
• Investigate molecular trends in coagulation, immune signaling, and metabolic dysfunction across control, CSKP, and CRKP patient groups.

Strengths

• Includes multi-omics data with 85 differentially expressed proteins and 128 differential metabolites identified.
• Features a four-protein biomarker panel independently validated in a separate cohort with a reported test AUC of 0.920.
• Data is openly shared under a CC-BY-4.0 license, facilitating reuse and redistribution.

Limitations

• 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.
• The dataset is small at 406.8 KB, indicating a limited sample scope.

Provenance

Source

figshare

Collection Method

Serum samples from sepsis patients underwent in-depth proteomic and metabolomic profiling, with machine learning applied to construct diagnostic models.

Freshness

Last updated 2026-05-20 05:44:19; freshness should be verified.