ProtPep1 is comprised of 182 complexes and was designed to exclude peptides involved in crystallographic contacts.We surveyed published protein-peptide complex datasets and found that many peptides in these complexes are involved in crystal contacts. These types of crystal contacts are more likely to occur with peptides which are often larger than drug–like ligands and tend to interact with proteins at shallow binding sites. This dataset of 182 complexes contains high resolution crystallographic structures with peptides ranging from 3 to 13 amino acids in length, and interacting solely with the biolog-ical unit of the protein they bind to. We believe that this dataset is more suitable for developing, cali-brating, and benchmarking computational methods related to protein peptide interaction.
Dataset is available here.