An immunoinformatic strategy for developing peptide vaccines against autoimmune diseases by targeting cross-reactive bacterial antigens.

Objective: To identify potential immunogenic peptide vaccine candidates targeting bacterial proteins implicated in autoimmune disorders via molecular mimicry, employing a population-specific immunoinformatic approach. Methodology: This cross sectional in silico bioinformatics study was carried out in Department of Immunology, Armed Forces Institute of Pathology, Rawalpindi, in August 2025. Fourteen proteins from eleven bacteria were selected due to their reported association with autoimmune disorders. FASTA sequences were retrieved and 15mer peptides generated. Binding affinities to three commonest HLA II alleles in Pakistan were determined using NetMHCIIpan-4.0. Strong binders were screened with BLASTP to remove 100% homologous peptides to human proteins. Rest of the peptides were assessed for CD4 T lymphocytes immunogenicity using Immune Epitope Database combined method. Results: From 14 bacterial proteins of 11 bacteria, 240 initial strong binders to HLA DRB1*03:01, 13:01 and 14:04 were identified. Four were excluded due to complete homology to at least one of human proteins. Of the remaining 236, 38 peptides exhibited high CD4 immunogenicity score (median percentile rank < 20). These peptides originated from eight proteins of seven bacterial species, each associated with specific autoimmune disorders. Conclusion: The immunoinformatic approach successfully identified 38 peptides as potential candidates for synthetic peptide vaccines against bacterial proteins implicated in autoimmunity. It has advantage of being rapid, population specific and reduces cost. It should be kept in mind that all bioinformatic findings must be confirmed by in vitro functional assays and in vivo studies before clinical trials.