A Minimal Computational Workflow for CDR-H3-Inspired Cyclic Peptide Design Targeting an Antibody-Defined Epitope

Abstract

Antibody therapeutics can effectively target protein–protein interactions; however, translating antibody recognition motifs into short peptide ligands remains challenging due to the loss of conformational control outside the antibody framework. In particular, CDR-derived peptides often require additional constraints to approximate antibody-defined binding geometries. Here, we present a minimal computational workflow for the design and filtering of short cyclic peptides inspired by an antibody CDR-H3 motif, using the IgE Fc region as a model antibody-defined epitope. A CDR-H3–derived sequence motif was extracted from a crystallographic IgE–anti-IgE antibody complex (PDB: 5HYS) and used to generate a small panel of disulfide-constrained cyclic peptide variants. Docking simulations were performed under a standardized protocol, and candidate peptides were compared based on pose reproducibility and spatial overlap with the antibody-defined epitope rather than absolute affinity prediction. Within this framework, one cyclic peptide variant consistently adopted binding poses overlapping the antibody-defined epitope and exhibited more stable docking behavior than a linear reference peptide. Although limited to computational analysis, this study demonstrates that antibody-derived CDR-H3 sequence information combined with minimal conformational constraints can bias short peptides toward biologically relevant epitope regions and provides a reproducible strategy for early-stage cyclic peptide design prior to experimental validation.