Transcriptome integration analysis of shared biomarkers and common immune mechanisms in SLE and PSO

This study aimed to identify shared diagnostic biomarkers and common immune mechanisms between systemic lupus erythematosus (SLE) and psoriasis (PSO) via integrated transcriptomic analysis, and to elucidate the role of genetic susceptibility in driving disease pathogenesis following viral infection. GEO datasets of SLE and PSO were analyzed. Shared genes were screened using differential expression analysis and WGCNA. 92 DEGs were identified, and 7 key shared genes ( OASL, SAMD9, IFI6, OAS3, NMI, UBE2L6, MX1 ) were determined after WGCNA and intersection analysis. Among 8 machine learning models, LASSO performed best: for SLE, training set AUC was 0.935, external validation AUCs were 0.764 (accuracy 0.745) and 0.844 (accuracy 0.896); for PSO, training set AUC was 0.841, internal validation AUC 0.910 (accuracy 0.989), external validation AUCs 0.941 (accuracy 0.968) and 0.869 (accuracy 0.815). Immune infiltration analysis showed significant correlations between key genes and specific immune cell subsets.Immunofluorescence analysis confirmed elevated protein expression levels of UBE2L6 and SAMD9 in both diseases.Single-cell analysis revealed that most key genes were differentially expressed in dendritic cells and monocytes in SLE, but in T cells in PSO. SLE and PSO share 7 susceptibility genes that are significantly enriched in immune response pathways related to viral infections. The genetic susceptibility of these genes can lead to imbalance or excessive activation of the body's antiviral defense, and through dysregulation of innate and adaptive immunity, promote the occurrence and development of the diseases. The LASSO model further supports the reliability of these genes as potential diagnostic biomarkers for SLE and PSO. • The 7 immune-related genes ( OASL, SAMD9, IFI6, OAS3, NMI, UBE2L6, MX1 ) identified through transcriptome analysis can serve as potential biomarkers for the co-occurrence of SLE and PSO. • Single-cell analysis elucidates that the key genes mainly cause SLE by regulating DC and monocytes, while in PSO, the key genes may drive disease progression by regulating T cell activity. • Through a comparative analysis of 8 machine learning algorithms, the diagnostic model based on LASSO performed the best, demonstrating strong potential for clinical application.