GraphNOSE: A Graph Transformer in Olfaction

Predicting olfactory qualities from molecular structure is an open problem in chemo-informatics. Although linear models can link molecular features to odor descriptors, they often fail when extrapolating to novel chemical scaffolds, extreme molecular weights, or complex odor mixtures. To address this, we introduce GraphNOSE, an open-source graph transformer framework that predicts multi-label odor descriptors for single molecules and binary mixtures. By integrating positional and structural encodings within a transformer-based graph architecture, GraphNOSE achieves strong performance with six times fewer parameters than standard graph neural network (GNN) baselines while consistently outperforming molecular foundational model embeddings, molecular fingerprints, and baseline GNNs by an average AUROC margin of ~ 4.5%. Finally, we apply explainable AI methods that validate which substructures and molecular features that drive odor predictions. Together, these results establish GraphNOSE as a scalable architecture for olfactory prediction that generalizes to compounds underrepresented in current perceptual databases. The code will be publicly available upon publication at https://github.com/CSIO-FPIL/GraphNOSE.