Olfactory Label Prediction on Aroma-Chemical Pairs

Laura Sisson,Aryan Amit Barsainyan,Mrityunjay Sharma,Ritesh Kumar

2024-06-05

Abstract:The application of deep learning techniques on aroma-chemicals has resulted in models more accurate than human experts at predicting olfactory qualities. However, public research in this domain has been limited to predicting the qualities of single molecules, whereas in industry applications, perfumers and food scientists are often concerned with blends of many molecules. In this paper, we apply both existing and novel approaches to a dataset we gathered consisting of labeled pairs of molecules. We present graph neural network models capable of accurately predicting the odor qualities arising from blends of aroma-chemicals, with an analysis of how variations in architecture can lead to significant differences in predictive power.

Machine Learning,Chemical Physics,Quantitative Methods

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve This paper aims to address the following issues: 1. **Odor Prediction of Aromatic Chemical Mixtures**: Existing research primarily focuses on odor prediction of individual molecules. However, in practical applications (such as perfumery and food science), it is often necessary to predict the odor of mixtures of multiple molecules. Therefore, the goal of the paper is to develop a model that can accurately predict the odor of mixtures of multiple aromatic chemicals. 2. **Application of Graph Neural Networks in Aromatic Chemical Mixtures**: The paper adopts existing Graph Neural Network (GNN) methods and proposes some new methods to generate embedding representations of mixtures. Through these models, the researchers aim to explore the impact of different architectures on predictive capabilities. 3. **Dataset Construction and Model Validation**: The paper constructs a dataset containing labeled mixture pairs and uses this dataset to train various GNN models, including Graph Isomorphism Network (GIN) and Message Passing Neural Network (MPNN). The performance of these models on the mixture prediction task is experimentally validated, and their transferability to single molecule prediction tasks is explored. In summary, this paper aims to solve the complex problem of odor prediction for multi-molecule mixtures using deep learning techniques, particularly Graph Neural Networks, and provides new methods and datasets to advance further research in this field.

Olfactory Label Prediction on Aroma-Chemical Pairs

Odor Classification: Exploring Feature Performance and Imbalanced Data Learning Techniques

Predicting odor from vibrational spectra: a data-driven approach

Insight into the Structure–Odor Relationship of Molecules: A Computational Study Based on Deep Learning

An Olfactory Sensor Array for Predicting Chemical Odor Characteristics from Mass Spectra with Deep Learning

Olfactory Perception Prediction Model Inspired by Olfactory Lateral Inhibition and Deep Feature Combination.

A deep position-encoding model for predicting olfactory perception from molecular structures and electrostatics

Predicting human olfactory perception from chemical features of odor molecules

Mol-PECO: a deep learning model to predict human olfactory perception from molecular structures

A framework for predicting odor threshold values of perfumes by scientific machine learning and transfer learning

Odor Prediction and Aroma Mixture Design Using Machine Learning Model and Molecular Surface Charge Density Profiles

An Odor Labeling Convolutional Encoder-Decoder for Odor Sensing in Machine Olfaction

DeepNose: An Equivariant Convolutional Neural Network Predictive Of Human Olfactory Percepts

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

Decoding Structure–Odor Relationship Based on Hypergraph Neural Network and Deep Attentional Factorization Machine

Multi-Label Code Smell Detection with Hybrid Model Based on Deep Learning

Harnessing Graph Neural Networks to Craft Fragrances Based on Consumer Feedback

Predicting individual perceptual scent impression from imbalanced dataset using mass spectrum of odorant molecules

Learning to smell for wellness

Classification of Tastants: A Deep Learning Based Approach

Molecule Generation and Optimization for Efficient Fragrance Creation