Material-Efficient Optimization of Peptide Bioconjugation via Sub-Micromole Automated Flow and Bayesian Optimization

We report a miniaturized, low-consumption optimization workflow for peptide bioconjugation by integrating automated continuous-flow experimentation with Bayesian optimization. Peptide substrates are often available only in limited quantities and exhibit substantial structural diversity, necessitating extensive condition screening and rendering conventional approaches inefficient. In this study, Trp-selective bioconjugation of leuprorelin was employed as a model reaction. The flow platform enabled precise control of reaction parameters and automated screening using minimal sample quantities (<1 μmol per trial). Bayesian optimization efficiently identified optimal conditions while significantly reducing experimental burden. The developed strategy achieved up to 82% yield with high reproducibility and minimized byproduct formation, demonstrating a practical approach for data-driven optimization of substratedependent peptide reactions involving limited and valuable materials.