PepMLM allows the design of peptide binders taking into account the target sequence only, not requiring the target structure. ProteinMPNN (Protein Message Passing Neural Network) is a deep learning-based model that rapidly and accurately predicts amino acid sequences for a given protein backbone structure. Developed by the Baker Lab, this inverse folding model uses a graph neural network to generate novel sequences that are highly likely to fold into the desired 3D shape. PepMLM was used to design peptide binders, followed by structure prediction with Boltz2. When the metrics (e.g., pLDDT) were below threshold, proteinMPNN was used to design sequences that fold in the target structure. Boltz2 was used to predict the complex structures.