Protein Complex Structure Prediction With AlphaFold-Enhanced HDOCK in CAPRI Rounds 47-55

Protein-protein interactions play a critical role in numerous biological processes, and understanding these interactions is essential for deciphering cellular mechanisms and designing therapeutic interventions. Predicting protein-protein complex structures by computational methods is an important approach to studying protein-protein interactions. The CAPRI (Critical Assessment of PRediction of Interactions) experiment has served as a benchmark for evaluating computational methods for predicting protein complex structures. We participated in CAPRI Rounds 47-55 and continuously refined our complex structure prediction strategies throughout this period. Initially, our approach was based on a hybrid docking strategy that combined template-based and ab initio docking methods. However, starting from Round 53, we integrated AlphaFold into our prediction pipeline. Inspired by the experiences of other participants in Round 54, we further refined our use of AlphaFold by enhancing the sampling strategy, which significantly improved our prediction accuracy in Round 55.