A Novel Framework for Multi-Modal Protein Representation Learning
Runjie Zheng, Zhen Wang, Anjie Qiao, Jiancong Xie, Jiahua Rao, Yuedong Yang
TL;DR
DAMPE tackles protein function prediction by unifying intrinsic sequence/structure with extrinsic PPI-GO context. It first aligns structure into the sequence space via Optimal Transport, then fuses modalities with a conditional diffusion-based CGG that denoises a heterogeneous graph conditioned on fused intrinsic features and GO priors. Theoretical analysis shows the CGG objective promotes the condition encoder to absorb graph-aware knowledge into protein representations, and empirical results demonstrate improved GO-term prediction (notably MF/BP) with favorable efficiency compared to GNN baselines. The framework achieves robust performance across GO branches, with ablations confirming the critical roles of OT-based alignment and CGG fusion, making it a scalable, theoretically grounded approach for multi-modal protein representation learning.
Abstract
Accurate protein function prediction requires integrating heterogeneous intrinsic signals (e.g., sequence and structure) with noisy extrinsic contexts (e.g., protein-protein interactions and GO term annotations). However, two key challenges hinder effective fusion: (i) cross-modal distributional mismatch among embeddings produced by pre-trained intrinsic encoders, and (ii) noisy relational graphs of extrinsic data that degrade GNN-based information aggregation. We propose Diffused and Aligned Multi-modal Protein Embedding (DAMPE), a unified framework that addresses these through two core mechanisms. First, we propose Optimal Transport (OT)-based representation alignment that establishes correspondence between intrinsic embedding spaces of different modalities, effectively mitigating cross-modal heterogeneity. Second, we develop a Conditional Graph Generation (CGG)-based information fusion method, where a condition encoder fuses the aligned intrinsic embeddings to provide informative cues for graph reconstruction. Meanwhile, our theoretical analysis implies that the CGG objective drives this condition encoder to absorb graph-aware knowledge into its produced protein representations. Empirically, DAMPE outperforms or matches state-of-the-art methods such as DPFunc on standard GO benchmarks, achieving AUPR gains of 0.002-0.013 pp and Fmax gains 0.004-0.007 pp. Ablation studies further show that OT-based alignment contributes 0.043-0.064 pp AUPR, while CGG-based fusion adds 0.005-0.111 pp Fmax. Overall, DAMPE offers a scalable and theoretically grounded approach for robust multi-modal protein representation learning, substantially enhancing protein function prediction.