Mapping the combinatorial coding between olfactory receptors and perception with deep learning скачать в хорошем качестве

Mapping the combinatorial coding between olfactory receptors and perception with deep learning

At the core of our sense of smell is olfactory information flow, in which odorant molecules activate a subset of our olfactory receptors (ORs) and combinations of unique receptor activations code for unique odors. Understanding this relationship is crucial for unraveling the mysteries of human olfaction and its potential therapeutic applications. We develop a novel, biologically-inspired approach that first maps odorant molecules to their respective OR activation profiles using geometric learning and protein language models, and subsequently predicts their odor percepts. Despite a lack of overlap between molecules with OR activation data and percept annotations, our joint model improves perception modeling by leveraging the receptor activation profile of each odorant as auxiliary features in predicting its percepts. In doing so, we hope to advance our understanding of receptor-mediated olfactory perception and the design of new odorants with desired perceptual qualities.