Constructing a disease-specific ceRNA coregulatory network for keratoconus diagnosis and landscape of the immune...

The early diagnosis of keratoconus (KC) is crucial for making treatment decisions. Therefore, the purpose of this study was to determine the potential disease-specific gene biomarker and landscape the immune environment in KC.

The transcriptome data of KC was obtained from Gene Expression Omnibus (GEO) and ArrayExpress databases for next analysis. The differently expressed mRNAs, microRNAs and lncRNAs between KC and control groups were firstly identified and the disease-specific protein-protein interaction (PPI) network as well as competing endogenous RNA (ceRNA) coregulatory network were constructed to explore the underlying molecular mechanism of KC. Besides, ElasticNet algorithm was used to develop a diagnostic model and associated nomograms to improve diagnosis of KC. Finally, multiple deconvolution methodologies were applied to decode the immune environment of KC patients.

In brief, we constructed the disease-specific PPI and ceRNA networks in KC through integrative analyses. The pathway enrichment manifested that these networks were significantly associated with lipopolysaccharide, chemokine and inflammatory related pathways. Based on the ceRNA network, we constructed a diagnostic model and associated nomogram which manifested a good performance for diagnosis of KC. Moreover, contrasted to control groups, we obviously observed that a distinct immune microenvironment existed in KC patients. Via single-cell sequencing analysis, we found that immune cells (Monocytes, Macrophages, and T cells) were strongly connected with corneal cells in KC patients.

In sum, we systematically constructed a diagnostic model and associated nomogram which provided novel biomarkers for the early detection of KC. Besides, our study comprehensively displayed the immune microenvironment of KC which provided new insights for understanding the molecular mechanism of KC.