A transcriptomics-based meta-analysis identifies a cross-tissue signature for sarcoidosis

Area of Science:

• Genomics
• Immunology
• Systems Biology

Background:

• Sarcoidosis is a complex granulomatous disease with an unknown cause, primarily involving a T-helper 1 (Th1) immune response.
• Transcriptome-wide expression studies have advanced the understanding of sarcoidosis pathogenesis.

Purpose of the Study:

• To conduct a cross-tissue, cross-platform meta-analysis of transcriptomic data to identify a robust gene expression signature for sarcoidosis.
• To develop and validate a diagnostic gene classifier for sarcoidosis.

Main Methods:

• Systematic search of Gene Expression Omnibus (GEO) database for transcriptomic data from blood and affected tissues.
• Meta-analysis of datasets, separated into training and testing cohorts, to identify differentially expressed genes.
• Pathway enrichment analysis to understand biological processes associated with sarcoidosis.

Main Results:

• Identified 690 differentially expressed genes across various tissues; 29 genes showed robust association with sarcoidosis in both blood and lung tissues.
• Key genes identified include LINC01278, GBP5, and PSMB9.
• A 16-gene classifier demonstrated high performance (AUC 0.711-0.964) in independent validation, with enriched pathways including IFN-γ, IL-1, IL-18, autophagy, and viral response.

Conclusions:

• This meta-analysis provides a comprehensive view of sarcoidosis expression profiles across different tissues.
• A novel 16-gene diagnostic classifier for sarcoidosis has been developed, offering a potential non-invasive tool to aid clinical diagnosis.