From Molecules to Mixtures: Learning Representations of Olfactory Mixture Similarity using Inductive Biases

TL;DR

This paper introduces POMMix, a hierarchical deep learning model that extends the principal odor map to represent and predict the similarity of complex olfactory mixtures, advancing digital olfaction understanding.

Contribution

POMMix is the first model to effectively represent and predict olfactory mixture similarity using hierarchical neural architectures and domain-specific inductive biases.

Findings

Achieves state-of-the-art performance on multiple datasets

Demonstrates good generalization to unseen molecules and mixture sizes

Highlights the synergy of domain knowledge and deep learning in low-data regimes

Abstract

Olfaction -- how molecules are perceived as odors to humans -- remains poorly understood. Recently, the principal odor map (POM) was introduced to digitize the olfactory properties of single compounds. However, smells in real life are not pure single molecules, but complex mixtures of molecules, whose representations remain relatively under-explored. In this work, we introduce POMMix, an extension of the POM to represent mixtures. Our representation builds upon the symmetries of the problem space in a hierarchical manner: (1) graph neural networks for building molecular embeddings, (2) attention mechanisms for aggregating molecular representations into mixture representations, and (3) cosine prediction heads to encode olfactory perceptual distance in the mixture embedding space. POMMix achieves state-of-the-art predictive performance across multiple datasets. We also evaluate the generalizability of the representation on multiple splits when applied to unseen molecules and mixture sizes. Our work advances the effort to digitize olfaction, and highlights the synergy of domain expertise and deep learning in crafting expressive representations in low-data regimes.