APCyc: Property-Informed Design of Cyclic Peptides via Automated Cyclization
Researchers have developed APCyc, a new framework for designing cyclic peptides with therapeutic potential. This target-aware system explicitly models the cyclization process and optimizes multiple drug-relevant properties simultaneously. APCyc addresses limitations of existing generative models by learning cyclization-aware representations and using Bayesian guidance to generate peptides that meet specific property objectives. AI
IMPACT Introduces a novel AI approach for designing therapeutic cyclic peptides with multi-property optimization.