From

Background: Rheumatoid arthritis (RA) is a chronic autoimmune disease in which dysregulated interleukin-6 (IL-6) signaling through the IL-6 receptor (IL-6R) plays a central pathogenic role. Although monoclonal antibodies targeting this pathway are clinically effective, their use is limited by parenteral administration, high cost, and systemic immunosuppression. Peptide-based inhibitors represent a complementary strategy for modulating cytokine-receptor interactions, offering advantages in design flexibility and manufacturability. Recent advances in immunoinformatics and artificial intelligence (AI) facilitate the rational identification of peptide candidates with favorable safety profiles. Methods: We developed an integrated AI-assisted discovery pipeline incorporating immunoinformatics-based safety screening (toxicity, allergenicity, and antigenicity prediction), structural modeling, molecular docking with MM/GBSA rescoring, and 100-ns molecular dynamics (MD) simulations to identify IL-6R-targeting peptides. Seven candidate peptides (P01-P07) were prioritized based on predicted safety, binding energetics, and structural stability. Lead candidates were experimentally evaluated using competitive ELISA assays for IL-6/IL-6R binding and cell-based assays measuring IL-6-induced STAT3 phosphorylation. Results: Computational analyses consistently identified peptide P01 as the top-ranked candidate, exhibiting stable binding conformations, persistent hydrogen bonding at the IL-6R interface, low RMSD and RMSF values during MD simulations, and the most favorable MM/GBSA binding free energy. Conclusion: This study establishes a safety-focused, AI-driven peptide discovery framework integrating in silico prioritization with experimental validation. Peptide P01 represents a tractable early-stage IL-6R antagonist that provides a foundation for future structure-guided optimization and development as a complementary therapeutic modality for IL-6-driven inflammatory diseases.