Abstract
Alzheimer's disease (AD), the leading cause of dementia, is a progressive neurodegenerative disorder marked by memory loss, cognitive decline, and characteristic neuropathology involving amyloid-β (Aβ) plaques and tau tangles. Among emerging therapeutic strategies, Aβ-targeted immunotherapy using monoclonal antibodies or peptide vaccines offers the most promising disease-modifying potential. Mimotopes, short peptides that mimic antigenic epitopes of Aβ, have recently gained attention as safe and effective candidates for vaccine development. In this study, we employed a computational immunoinformatics approach to identify novel Aβ(42)-mimicking peptides derived from non-coding DNA sequences, representing an unconventional yet rich source of bioactive peptides. A virtual peptide library was generated from intergenic regions of the Escherichia coli genome and screened using B-cell epitope prediction, MHC-binding analysis, and structural similarity modeling to identify potential immunogenic mimotopes. Selected candidates were further evaluated through peptide-antibody docking with Aβ(42)-specific antibody fragments to assess binding affinity and epitope mimicry. Our findings demonstrate a novel computational framework for mining non-coding DNA to identify therapeutic peptide mimotopes. The identified Aβ(42)-like peptides exhibit strong potential as synthetic vaccine candidates for Alzheimer's disease, supporting a new direction in rational vaccine design against neurodegenerative disorders.