Table 3b.
Model results.
| Name | Acc | NER | F1 | Pr | ROC AUC | PR AUC | R | MCC | Sn | Sp | R2 | RMSE |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SMILES to Smell: Decoding the Structure–Odor Relationship of Chemical Compounds Using the Deep Neural Network Approach (Sharma et al., 2021b) | 0.98 | – | 0.88 | 0.97 | micro avg: 0.87 macro avg: 0.7 | micro avg: 0.74 macro avg: 0.75 | – | – | 0.84 | – | – | – |
| Predicting odor from vibrational spectra: a data-driven approach (Ameta et al., 2024) | – | – | 0.41 | 0.36 | – | – | – | – | 0.47 | – | – | – |
| Predicting natural language descriptions of mono-molecular odorants (Gutiérrez et al., 2018) | – | – | – | – | 0.66 | – | – | – | – | – | – | – |
| A principal odor map unifies diverse tasks in olfactory perception (Lee et al., 2023) | – | – | – | – | 0.89∗ | – | – | – | – | – | – | – |
| Odor Impression Prediction from Mass Spectra (Nozaki and Nakamoto, 2016) | – | – | – | – | – | – | 0.76∗ | – | – | – | – | – |
| Application of artificial intelligence to decode the relationships between smell, olfactory receptors and small molecules (Achebouche et al., 2022) | ||||||||||||
| odor note: | – | – | – | – | 0.82∗ | 0.24∗ | – | – | – | – | – | – |
| odor group: | – | – | – | – | 0.80∗ | 0.40∗ | – | – | – | – | – | – |
| XGBoost odor prediction model: finding the structure-odor relationship of odorant molecules using the extreme gradient boosting algorithm (Tyagi et al., 2024) | 0.99 | – | 0.99 | 0.99 | 0.99 | 0.99 | – | 0.99 | 0.99 | – | – | – |
| Predictive modeling for odor character of a chemical using machine learning combined with natural language processing (Nozaki and Nakamoto, 2018) | 0.69∗ | – | 0.63∗ | 0.78∗ | – | – | – | 0.40∗ | 0.53∗ | 0.85∗ | – | – |
| iBitter-SCM: Identification and characterization of bitter peptides using a scoring card method with propensity scores of dipeptides (Charoenkwan et al., 2020a) | 0.84 | – | – | – | 0.90 | – | – | 0.69 | 0.84 | 0.84 | – | – |
| e-Bitter: Bitterant Prediction by the Consensus Voting From the Machine-Learning Methods (Zheng et al., 2018) | 0.93 | – | 0.94 | 0.92 | – | – | – | 0.86 | 0.95 | 0.90 | – | – |
| BitterX: a tool for understanding bitter taste in humans (Huang et al., 2016) | 0.92 | – | – | 0.91 | 0.96 | – | – | – | 0.94 | 0.91 | – | – |
| Bitter or not? BitterPredict, a tool for predicting taste from chemical structure (Dagan-Wiener et al., 2017) | 0.83 | – | – | – | – | – | – | – | 0.77 | 0.86 | – | – |
| BERT4Bitter: a bidirectional encoder representations from transformers (BERT)-based model for improving the prediction of bitter peptides (Charoenkwan et al., 2021a) | 0.92 | – | – | – | 0.96 | – | – | 0.84 | 0.94 | 0.91 | – | – |
| iBitter-Fuse: A Novel Sequence-Based Bitter Peptide Predictor by Fusing Multi-View Features (Charoenkwan et al., 2021c) | 0.93 | – | – | – | 0.93 | – | – | 0.86 | 0.94 | 0.92 | – | – |
| Prediction of bitterant and sweetener using structure-taste relationship models based on an artificial neural network (Bo et al., 2022) | 0.88 | – | – | 0.89 | 0.95 | – | – | 0.75 | 0.83 | 0.93 | – | – |
| Novel scaffold of natural compound eliciting sweet taste revealed by machine learning (Bouysset et al., 2020) | – | – | – | – | – | – | – | – | – | – | 0.75 | – |
| Sweetness prediction of natural compounds (Chéron et al., 2017b) | – | – | – | – | – | – | – | – | – | – | 0.85 | – |
| A QSTR-Based Expert System to Predict Sweetness of Molecules (Rojas et al., 2017) | – | 0.85 | – | – | – | – | – | – | 0.88 | 0.82 | – | – |
| Machine learning models to predict sweetness of molecules (Goel et al., 2023) | – | – | – | – | – | – | 0.94 | – | – | – | – | 0.33 |
| VirtualTaste: a web server for the prediction of organoleptic properties of chemical compounds (Fritz et al., 2021) | ||||||||||||
| sweet: | 0.89 | – | 0.88 | – | 0.95 | – | – | – | 0.68 | 0.92 | – | – |
| bitter: | 0.90 | – | 0.88 | – | 0.96 | – | – | – | 0.88 | 0.97 | – | – |
| sour: | 0.97 | – | 0.84 | – | 0.99 | – | – | – | 0.80 | 0.99 | – | – |
| Predicting multiple taste sensations with a multiobjective machine learning method (Androutsos et al., 2024) | 0.72 | – | 0.74 | 0.79 | 0.87 | – | – | – | 0.72 | – | – | – |
| BitterSweetForest: A Random Forest Based Binary Classifier to Predict Bitterness and Sweetness of Chemical Compounds (Banerjee and Preissner, 2018) | ||||||||||||
| sweet | 0.97 | – | 0.92 | – | 0.98 | – | – | – | 0.91 | 0.97 | – | – |
| bitter | 0.97 | – | 0.95 | – | 0.98 | – | – | – | 0.97 | 0.91 | – | – |
| Informed classification of sweeteners/bitterants compounds via explainable machine learning (Maroni et al., 2022) | – | – | 0.88∗ | 0.87∗ | 0.95∗ | 0.94∗ | – | – | 0.89∗ | – | – | – |
| BitterSweet: Building machine learning models for predicting the bitter and sweet taste of small molecules (Tuwani et al., 2019) | ||||||||||||
| sweet | – | 0.83 | 0.86 | – | 0.88 | 0.95 | – | – | 0.79 | 0.88 | – | – |
| bitter | – | 0.82 | 0.86 | – | 0.88 | 0.93 | – | – | 0.85 | 0.79 | – | – |
| e-Sweet: A Machine-Learning Based Platform for the Prediction of Sweetener and Its Relative Sweetness (Zheng et al., 2019) | 0.91 | 0.9 | 0.88 | 0.9 | – | – | – | 0.81 | 0.86 | 0.94 | 0.78 | 0.28 |
| A Comprehensive Comparative Analysis of Deep Learning Based Feature Representations for Molecular Taste Prediction (Song et al., 2023) | ||||||||||||
| sweet | 0.90 | – | 0.85 | 0.84 | – | – | – | – | 0.86 | 0.92 | – | – |
| bitter | 0.89 | – | 0.87 | 0.93 | – | – | – | – | 0.82 | 0.95 | – | – |
| umami | 0.99 | – | 0.91 | 1.00 | – | – | – | – | 0.84 | 1.00 | – | – |
| BoostSweet: Learning molecular perceptual representations of sweeteners (Lee et al., 2022) | – | 0.90 | 0.91 | – | 0.96 | 0.97 | – | – | 0.91 | 0.90 | – | – |
| A novel multi-layer prediction approach for sweetness evaluation based on systematic machine learning modeling (Yang et al., 2022) | – | – | – | – | – | – | – | – | – | – | 0.85 | 0.48 |
| Toward a general and interpretable umami taste predictor using a multi-objective machine learning approach (Pallante et al., 2022) | 0.88 | – | 0.79 | – | 0.85 | – | – | – | 0.79 | 0.92 | – | – |
| iUmami-SCM: A Novel Sequence-Based Predictor for Prediction and Analysis of Umami Peptides Using a Scoring Card Method with Propensity Scores of Dipeptides (Charoenkwan et al., 2020b) | 0.87 | – | – | – | 0.90 | – | – | 0.68 | 0.71 | 0.93 | – | – |
| UMPred-FRL: A New Approach for Accurate Prediction of Umami Peptides Using Feature Representation Learning (Charoenkwan et al., 2021b) | 0.82 | – | – | – | 0.91 | – | – | 0.56 | 0.57 | 0.93 | – | – |
| Umami-BERT: An interpretable BERT-based model for umami peptides prediction (Zhang et al., 2023) | 0.95∗ | – | – | – | 0.97∗ | – | – | 0.85∗ | 0.97∗ | 1.00∗ | – | – |
| Umami-MRNN: Deep learning-based prediction of umami peptide using RNN and MLP (Qi et al., 2023) | 0.93 | – | – | – | 0.97 | – | – | 0.86 | 0.97 | 0.91 | – | – |
| Artificial neural networks modeling of non-fat yogurt texture properties: effect of process conditions and food composition (Batista et al., 2021) | ||||||||||||
| firmness | – | – | – | – | – | – | – | – | – | – | 1.00∗∗ | 0.00∗∗ |
| cohesiveness | – | – | – | – | – | – | – | – | – | – | 1.00∗∗ | 0.00∗∗ |
| adhesiveness | – | – | – | – | – | – | – | – | – | – | 0.98∗∗ | 0.00∗∗ |
| gumminess | – | – | – | – | – | – | – | – | – | – | 1.00∗∗ | 0.00∗∗ |
| Predicting the Textural Properties of Plant-Based Meat Analogs with Machine Learning (Kircali Ata et al., 2023) | ||||||||||||
| hardness | – | – | – | – | – | – | – | – | – | – | – | 10.10∗∗ |
| chewiness | – | – | – | – | – | – | – | – | – | – | – | 6.04∗∗ |
| Rapid and Non-Invasive Assessment of Texture Profile Analysis of Common Carp (Cyprinus carpio L.) Using Hyperspectral Imaging and Machine Learning (Cao et al., 2023) | 32 targets, see paper for full results. | |||||||||||
| Prediction of textural characteristics in low-fat mozzarella cheese by Hyperspectral imaging using machine learning methods (Jahani et al., 2024) | ||||||||||||
| hardness | – | – | – | – | – | – | – | – | – | – | 0.88∗∗ | – |
| adhesiveness | – | – | – | – | – | – | – | – | – | – | 0.81∗∗ | – |
| springiness | – | – | – | – | – | – | – | – | – | – | 0.85∗∗ | – |
| cohesiveness | – | – | – | – | – | – | – | – | – | – | 0.70∗∗ | – |
| gumminess | – | – | – | – | – | – | – | – | – | – | 0.82∗∗ | – |
| chewiness | – | – | – | – | – | – | – | – | – | – | 0.84∗∗ | – |
| Non-destructive prediction of texture of frozen/thaw raw beef by Raman spectroscopy (Chen et al., 2020) | ||||||||||||
| tenderness | – | – | – | – | – | – | – | – | – | – | 0.81∗∗ | – |
| chewiness | – | – | – | – | – | – | – | – | – | – | 0.80∗∗ | – |
| firmness | – | – | – | – | – | – | – | – | – | – | 0.81∗∗ | – |
| hardness | – | – | – | – | – | – | – | – | – | – | 0.82∗∗ | – |
| springiness | – | – | – | – | – | – | – | – | – | – | 0.53∗∗ | – |
| Prediction of textural changes in grass carp fillets as affected by vacuum freeze drying using hyperspectral imaging based on integrated group wavelengths (Ma et al., 2017) | ||||||||||||
| WBSF | – | – | – | – | – | – | – | – | – | – | 0.91∗∗ | – |
| hardness | – | – | – | – | – | – | – | – | – | – | 0.90∗∗ | – |
| gumminess | – | – | – | – | – | – | – | – | – | – | 0.88∗ | – |
| chewiness | – | – | – | – | – | – | – | – | – | – | 0.90∗∗ | – |
| Prediction of color, texture, and sensory characteristics of beef steaks by visible and near infrared reflectance spectroscopy. A feasibility study (Liu et al., 2003) | 0.96∗∗ | – | – | – | – | – | – | – | – | – | – | – |
| Rapid determination of the textural properties of silver carp (Hypophthalmichthys molitrix) using near-infrared reflectance spectroscopy and chemometrics (Zhou et al., 2020) | ||||||||||||
| WHC | – | – | – | – | – | – | – | – | – | – | 0.95∗∗ | – |
| hardness | – | – | – | – | – | – | – | – | – | – | 0.89∗∗ | – |
| resilience | – | – | – | – | – | – | – | – | – | – | 0.87∗∗ | – |
| springiness | – | – | – | – | – | – | – | – | – | – | 0.83∗∗ | – |
| chewiness | – | – | – | – | – | – | – | – | – | – | 0.92∗∗ | – |
| shear force | – | – | – | – | – | – | – | – | – | – | 0.88∗∗ | – |
| Potential of hyperspectral imaging combined with chemometric analysis for assessing and visualising tenderness distribution in raw farmed salmon fillets (He et al., 2014) | ||||||||||||
| WBSF | – | – | – | – | – | – | – | – | – | – | 0.90 | – |
The performance of all models listed in Table 3a. Metrics include accuracy (Acc), non-error rate (NER), F1 score, precision (Pr), area under the receiver operating characteristic curve (ROC AUC), R score, Matthews correlation coefficient (MCC), sensitivity (Sn), specificity (Sp), R2 score, and root mean squared error (RMSE).