Shotgun Approach for PROTAC Ternary Complex Modeling and Evaluation

Proteolysis targeting chimeras (PROTACs) are bifunctional molecules that induce protein degradation by forming ternary complexes with the target protein (protein of interest, POI) and the E3 ubiquitin ligase. Most commonly used degrader design strategies start with known POI and E3 binders, followed by the replacement of different linkers. Structure-based linker design is promising, yet successful cases using ternary complex structures to guide linker design remain scarce, primarily due to the inherent flexibility and conformational diversity of the complexes. This challenge is compounded by two key limitations: first, obtaining high-quality POI-egrader-E3 ternary complex crystal structures faces major technical hurdles; second, existing crystal structures in the PDB only capture static snapshots of these dynamically changing systems, with some even biased by crystal packing. Although many computational tools for ternary complex modeling have been developed so far, their accuracy still needs improvement. Currently, the commonly used methods for evaluating ternary complex modeling all rely on comparison with crystal structures while ignoring potential issues that may exist. Therefore, it is necessary to develop new modeling and evaluation methods to guide degrader design. In this study, the conformation distribution extracted from multiple parallel MD trajectories for a ternary complex is used as a reference, and we propose a framework consisting of a new modeling protocol and a scoring to generate and evaluate a ternary complex ensemble. The results of 5 tested cases demonstrate that our protocol generates complexes that can cover or are at least close to the high-density regions of the MD trajectory. This framework is particularly suitable for the POI-degrader-E3 system, which has multiple possible stable conformations and PPIs.