Presentation Details
Comparing Odor Perception Across Humans and Mice

Elizabeth Hamel1, Robert Pellegrino1, Mebelyn Urena1, Sebastian Ceballo2, Saeed Karimimehr2, Khristina Samoilova3, Matt Andres1, Jenny Margolis1, Alexei Koulakov3, Dima Rinberg2, Joel Mainland1.

1Monell Center, Philadelphia, PA, USA.2New York University, New York City, NY, USA.3Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA

Abstract


Understanding the relationship between neural activity and perception is a critical question in neuroscience, particularly in the domain of olfaction where we are just beginning to develop methods to robustly quantify perceptual odor quality. Mice and humans have distinct advantages for relating neural activity to perception. Mouse models enable researchers to easily access neural activity, while humans allow researchers to more readily assess detailed perceptual reports. A critical step in establishing this bridge between species involves verifying how well perceptual similarity of odors observed in one species can predict perception in another species. Using a set of common stimuli, we measured the perceptual similarity of 20 odorants in mice using a delayed-match-to-sample (DMTS) task (Nakayama, 2022), and in humans using two separate methods; first the human equivalent to the DMTS, a two-alternative forced-choice paradigm (2AFC), and second, an odor profiling method. We found that, when compared to the 2AFC, perceptual similarity between the species was not significantly correlated (p = 0.13). In contrast, perceptual distances generated from the profiling task were weakly correlated with mouse perceptual distances (r = 0.31, p <0.03).  Both human methods were strongly correlated (r = 0.49, p <0.001). Further analysis, using multidimensional scaling, revealed that the first dimension of human perceptual space can be predicted using mouse perceptual embedding. As in previous research, this perceptual dimension is consistent with the perceived pleasantness of the odors. Examining additional odors will allows us to more densely sample perceptual spaces and determine which dimensions can be predicted across species.

No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.

View Questions & Answers for this Presentation

No questions or discussion have been posted for this presentation yet. You may ask a questions using the tab above. Please note that ONLY the questions the presenter chooses to answer publicly will be posted here.



Ask A Question

You may submit a question to the presenter below. All questions may not receive a response.


View/Download Proceedings Manuscript