Grafting-based design of cyclic peptides from nanobodies for the molecular recognition of protein targets

This thesis investigates the design of constrained cyclic peptides derived from nanobody CDR3 loops as potential molecular recognition elements for protein targets. Experimentally determined nanobody–antigen complexes were analyzed to identify CDR3 regions involved in antigen recognition. Selected CDR3 sequences were redesigned by introducing terminal cysteine residues to enable potential disulfide-mediated cyclization, and peptide–antigen complexes were modeled using AlphaFold 3. Structural comparison with the corresponding experimental nanobody–antigen complexes was performed using RMSD analysis and interface evaluation. The selected peptide candidates were further investigated through molecular dynamics simulations to assess their conformational stability, flexibility, and persistence of peptide–protein interactions. Scoring function approaches were also considered to provide additional insights into the relative binding behavior of the designed complexes. Overall, this work integrates computational structural analysis, peptide design, molecular modeling, and molecular dynamics simulations to evaluate the potential of nanobody-derived cyclic peptides as compact molecular recognition scaffolds.