Integrative transcriptomic and systems biology analysis of public SLE datasets identifies immune regulatory pathways

Systemic lupus erythematosus (SLE) is a multifactorial autoimmune disease characterised by loss of immune tolerance and chronic inflammation, but its molecular pathogenesis remains incompletely understood. In this work we examine whether immune regulatory transcripts and pathways are recurrently detectable across heterogeneous publicly available SLE transcriptomic datasets, and we explore one of the recurrent pathways with a kinetic systems-biology model. We analyzed one microarray dataset (E-GEOD-46923, profiled on Affymetrix HG-U133A and HG-U133B), two bulk RNA-seq datasets (E-MTAB-7145, E-MTAB-11919), and two single-cell RNA-seq datasets (GSE135779, GSE163121) using a unified differential-expression criterion (|log2 fold change|≥ 1, Benjamini-Hochberg adjusted p-value < 0.05). KEGG pathway enrichment was performed with a per-dataset background gene universe. A mass-action kinetic model of Th1/Th2 differentiation was constructed in MATLAB SimBiology, and global sensitivity analysis was performed using the variance-based Sobol method. Across the heterogeneous datasets, CD53, IFITM1, and RPL11 were recurrently identified as differentially expressed transcripts, and the Th1/Th2 cell differentiation pathway, together with related cytokine-cytokine receptor and JAK-STAT pathways, emerged as a recurrent immune-regulatory signal. Systems-biology simulation under SLE-derived initial conditions predicted atypical IL-2 and GATA3 expression dynamics, which is consistent with, but does not by itself prove, the cytokine-signalling dysregulation reported in SLE. Sobol sensitivity analysis identified IL-4 and the modelled co-stimulatory and Notch ligand species (CDN1-6 and Jagged1/2) as the largest non-additive regulators of IL-2 in the model. Overall, this work integrates transcriptomic recurrence analysis with kinetic modelling to generate testable hypotheses regarding immune regulatory dysfunction in SLE.