Classification-based machine learning approaches to predict the taste of molecules: A review

Rojas, Cristián Guerra; Ballabio, Davide; Consonni, Viviana; Suárez-Estrella, Diego; Todeschini, Roberto

doi:10.1016/j.foodres.2023.113036

Cited by 12 publications

(4 citation statements)

References 124 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The primary objective was to ascertain the applicability of deep learning-driven molecular representation techniques and to identify the optimal approaches for addressing taste prediction tasks, an aspect that has been notably absent in previous research endeavors. Since previous authors have summarized the performance of the taste prediction tools referred to in their original papers [ 39 , 40 ], we alternatively examined the performance of various methods with a consistent and comprehensive database. In addition, we employed a robust set of metrics to gauge their effectiveness.…”

Section: Discussionmentioning

confidence: 99%

A Comprehensive Comparative Analysis of Deep Learning Based Feature Representations for Molecular Taste Prediction

Song,

Chang,

Tian

et al. 2023

Foods

View full text Add to dashboard Cite

Taste determination in small molecules is critical in food chemistry but traditional experimental methods can be time-consuming. Consequently, computational techniques have emerged as valuable tools for this task. In this study, we explore taste prediction using various molecular feature representations and assess the performance of different machine learning algorithms on a dataset comprising 2601 molecules. The results reveal that GNN-based models outperform other approaches in taste prediction. Moreover, consensus models that combine diverse molecular representations demonstrate improved performance. Among these, the molecular fingerprints + GNN consensus model emerges as the top performer, highlighting the complementary strengths of GNNs and molecular fingerprints. These findings have significant implications for food chemistry research and related fields. By leveraging these computational approaches, taste prediction can be expedited, leading to advancements in understanding the relationship between molecular structure and taste perception in various food components and related compounds.

show abstract

Section: Discussionmentioning

confidence: 99%

A Comprehensive Comparative Analysis of Deep Learning Based Feature Representations for Molecular Taste Prediction

Song,

Chang,

Tian

et al. 2023

Foods

View full text Add to dashboard Cite

show abstract

“…Current research on taste is basically on single molecules. These taste molecules are able to present five basic tastes of sweet, sour, bitter, salty, and umami, and even some single-taste molecules with two or more tastes (Rojas et al, 2023). Besides, food is usually composed of multiple taste substances that can trigger different taste sensations simultaneously, and different taste molecules can interact each other.…”

Section: Crosstalk Between Tastesmentioning

confidence: 99%

Taste and its receptors in human physiology: A comprehensive look

Li,

Sun

et al. 2024

Food Frontiers

View full text Add to dashboard Cite

Increasing evidence shows that food has significance beyond traditional perception (providing nutrition and energy) in maintaining normal life activities. It is indicated that the sense of taste plays a crucial part in regulating human life activities. Taste is one of the basic physiological sensations in mammals, and it is the fundamental guarantee for them to perceive, select, and ingest nutrients in order to survive. With the advances in electrophysiology, molecular biology, and structural biology, studies on the intracellular and extracellular transduction mechanisms of taste have made great progress and gradually revealed the indispensable role of taste receptors in the regulation and maintenance of normal physiological activities. Up to now, how food regulates life activities through the taste pathway remains unclear. Thus, this review comprehensively and systematically summarizes the current study about the sense of taste, the function of taste receptors, the taste–structure relationship of gustatory molecules, the cross‐talking between distinctive tastes, and the role of the gut–organ axis in the realization of taste. Moreover, we also provide forward‐looking perspectives on taste research to afford a scientific basis for revealing the scientific connotation of taste receptors regulating body health.

show abstract

“…This disparity underscores the underexplored chemical space of antioxidant peptides, primarily due to the intricate nature of their identification process. To address this challenge, machine learning models offer a promising alternative to traditional methods for rapidly discovering compounds with desired properties, as demonstrated by recent advancements in using ML for anticancer and antimicrobial peptides and food chemistry 14 – 17 . Consequently, in this research, for the first time, we have developed a generative model based on recurrent neural networks (RNNs) for the de novo design of novel AOPs (Fig.…”

Section: Introductionmentioning

confidence: 99%

De novo antioxidant peptide design via machine learning and DFT studies

Hesamzadeh,

Seif,

Mahmoudzadeh

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Antioxidant peptides (AOPs) are highly valued in food and pharmaceutical industries due to their significant role in human function. This study introduces a novel approach to identifying robust AOPs using a deep generative model based on sequence representation. Through filtration with a deep-learning classification model and subsequent clustering via the Butina cluster algorithm, twelve peptides (GP1–GP12) with potential antioxidant capacity were predicted. Density functional theory (DFT) calculations guided the selection of six peptides for synthesis and biological experiments. Molecular orbital representations revealed that the HOMO for these peptides is primarily localized on the indole segment, underscoring its pivotal role in antioxidant activity. All six synthesized peptides exhibited antioxidant activity in the DPPH assay, while the hydroxyl radical test showed suboptimal results. A hemolysis assay confirmed the non-hemolytic nature of the generated peptides. Additionally, an in silico investigation explored the potential inhibitory interaction between the peptides and the Keap1 protein. Analysis revealed that ligands GP3, GP4, and GP12 induced significant structural changes in proteins, affecting their stability and flexibility. These findings highlight the capability of machine learning approaches in generating novel antioxidant peptides.

show abstract

Classification-based machine learning approaches to predict the taste of molecules: A review

Cited by 12 publications

References 124 publications

A Comprehensive Comparative Analysis of Deep Learning Based Feature Representations for Molecular Taste Prediction

A Comprehensive Comparative Analysis of Deep Learning Based Feature Representations for Molecular Taste Prediction

Taste and its receptors in human physiology: A comprehensive look

De novo antioxidant peptide design via machine learning and DFT studies

Contact Info

Product

Resources

About