Abstract
Nuclear magnetic resonance (NMR) spec-troscopy plays an essential role in deciphering molecular structure and dynamic behaviors. While AI-enhanced NMR prediction models hold promise, challenges still persist in tasks such as molecular retrieval, iso-mer recognition, and peak assignment. In response, this paper introduces a novel solution , Knowledge-Guided Multi-Level Multi-modal Alignment with Instance-Wise Discrimination (K-M 3 AID), which establishes correspondences between two heterogeneous modalities: molecular graphs and NMR spectra. K-M 3 AID employs a dual-coordinated contrastive learning architecture with three key modules: a graph-level alignment module, a node-level alignment module, and a communication channel. Notably, K-M 3 AID introduces knowledge-guided instance-wise discrimination into con-trastive learning within the node-level alignment module. In addition, K-M 3 AID demonstrates that skills acquired during node-level alignment have a positive impact on graph-level alignment , acknowledging meta-learning as an inherent property. Empirical validation underscores the effectiveness of K-M 3 AID in multiple zero-shot tasks.