AROMMA: Unifying Olfactory Embeddings for Single Molecules and Mixtures

TL;DR

AROMMA introduces a unified embedding framework for single molecules and mixtures, leveraging attention mechanisms and knowledge distillation to improve odor representation learning across diverse datasets.

Contribution

It is the first to unify odor embeddings for molecules and mixtures, using attention-based aggregation and knowledge distillation for enhanced generalization.

Findings

Achieves up to 19.1% AUROC improvement on benchmark datasets.

Unifies odor representations for molecules and mixtures.

Demonstrates robust generalization across domains.

Abstract

Public olfaction datasets are small and fragmented across single molecules and mixtures, limiting learning of generalizable odor representations. Recent works either learn single-molecule embeddings or address mixtures via similarity or pairwise label prediction, leaving representations separate and unaligned. In this work, we propose AROMMA, a framework that learns a unified embedding space for single molecules and two-molecule mixtures. Each molecule is encoded by a chemical foundation model and the mixtures are composed by an attention-based aggregator, ensuring both permutation invariance and asymmetric molecular interactions. We further align odor descriptor sets using knowledge distillation and class-aware pseudo-labeling to enrich missing mixture annotations. AROMMA achieves state-of-the-art performance in both single-molecule and molecule-pair datasets, with up to 19.1% AUROC improvement, demonstrating a robust generalization in two domains.