USP-ddG: A Unified Structural Paradigm with Data Efficacy and Mixture-of-Experts for Predicting Mutational Effects on Protein-Protein Interactions

Abstract

Accurately estimating changes in binding free energy ({Delta}{Delta}G) is essential for understanding protein-protein interactions (PPIs) and guiding rational protein design. Existing deep learning approaches benefit from pre-training on large structural datasets, but they often require heavy computation and overlook data efficacy. In addition, prior models typically assume fixed backbone structures and ignore conformational dynamics, limiting generalization. We present USP-ddG, a unified structural paradigm for{Delta}{Delta} G prediction. USP-ddG integrates a dual-channel architecture with three complementary components: (i) inverse folding-based log-odds ratio, (ii) empirical energy terms from FoldX, and (iii) a geometric encoder with Gaussian noise to capture relaxed conformations. To enhance representation power, we introduce a framework that integrates feed-forward network (FFN) and Mixture-of-Experts (MoE) to model domain-invariant and -specific features, respectively. We further propose CATH-guided Folding Ordering (CFO), a data efficacy strategy that organizes samples to mitigate catastrophic forgetting and data distribution bias. USP-ddG consistently outperforms existing state-of-the-art (SoTA) methods on the SKEMPI v2.0 benchmark, including the challenging hold-out CATH test set. It achieves superior accuracy on both single- and multi-point mutations and demonstrates strong performance in antibody affinity optimization against H1N1 and HER2, and in assessing SARS-CoV-2 variants to hACE2. Ablation studies confirm the benefit of each component. These results highlight USP-ddG as a robust and data-efficient framework for modeling mutational effects on PPIs. AvailabilityUSP-ddG is publicly available at https://github.com/ak422/USP-ddG.