Publication
Cyclic peptides are attractive for drug discovery due to their excellent binding properties and the potential to cross cell membranes. However, by far not all cyclic peptides are cell permeable, and measuring or predicting their membrane permeability is not trivial. In this work, we assessed the membrane permeability of thioether‐cyclized peptides, a widely used format in drug discovery. We developed a strategy for synthesizing hundreds of cyclic peptides carrying a short chloroalkane tag for the bulk quantification of membrane permeability in live cells using the chloroalkane penetration assay. Permeability data for random cyclic peptides established design rules, namely that the probability of peptides entering cells is strongly increasing if the molecular weight is below 800 Da, the polar surface is smaller than 250 Å2, or if there are less than six H‐bond donors. From this, machine learning could predict the membrane permeability of random peptides with good confidence, facilitating the future development of membrane permeable cyclic peptide drugs.