QSBR Study of Bitter Taste of Peptides: Application of GA-PLS in Combination with MLR, SVM, and ANN Approaches

Abstract

Detailed information about the relationships between structures and properties/activities of peptides as drugs and nutrients is useful in the development of drugs and functional foods containing peptides as active compounds. The bitterness of the peptides is an undesirable property which should be reduced during drug/nutrient production, and quantitative structure bitter taste relationship (QSBR) studies can help researchers to design less bitter peptides with higher target efficiency. Calculated structural parameters were used to develop three different QSBR models (i.e., multiple linear regression, support vector machine, and artificial neural network) to predict the bitterness of 229 peptides (containing 2–12 amino acids, obtained from the literature). The developed models were validated using internal and external validation methods, and the prediction errors were checked using mean percentage deviation and absolute average error values. All developed models predicted the activities successfully (with prediction errors less than experimental error values), whereas the prediction errors for nonlinear methods were less than those for linear methods. The selected structural descriptors successfully differentiated between bitter and nonbitter peptides.

Structural characteristics of food protein-originating di- and tripeptides using principal component analysis

“…First, they were the most numerous group of bitter peptides listed in BIOPEP-UWM database comparing to longer chain sequences [19] and such sets of di-and tripeptides enabled to obtain well-conditioned matrices [25]. According to scientific reports, chemometric models developed for each subset analyzed generate less predictability errors [27]. Second, many bitter di-and tripeptides tend to exhibit additional biological function, mainly enzyme inhibiting, antioxidative, and/or antithrombotic.…”

“…According to Soltani et al [27], hydrophobicity and an increasing number of carbon atoms in the side chains of amino acids forming peptides affect the bitterness of peptides. These attributes are related to the bulkiness and molecular weight of amino acids.…”

Valorisation of protein hydrolysates from animal by‐products: perspectives on bitter taste and debittering methods: a review

“…Recently, quantitative structure‐bitter taste relationship (QSBR) analysis has been performed by Soltani et al . (). According to multiple regression analysis, artificial neural networks and support vector machine, bitterness of peptides is intensified with an increased number of amino acids within a peptide chain (Soltani et al ., ).…”

“…As a supplement to sensory evaluations, researchers have developed QSAR (quantitative structure‐activity relationship) models to predict bitter intensity of peptides based on their structural characteristics. In the past years, different analytical methods have been developed to predict bitterness, including multiple linear regression (Kim & Li‐Chan, ), partial least square regression (Kim & Li‐Chan, ; Wu & Aluko, ), support vector regression (Soltani et al ., ), artificial neural network (Soltani et al ., ) and principal component regression (Liu et al ., ). Kim & Li‐Chan () employed QSAR as a tool to investigate the structural features of bitter peptides.…”

Prediction of the Health Effects of Food Peptides and Elucidation of the Mode‐of‐action Using Multi‐task Graph Convolutional Neural Network

“…ML methods have been used to analyze the quantitative structure activity/property relationships (QSAR/QSPR) [13][14][15][16] for bitter taste [17,18] and antioxidant. ML methods have been used to analyze the quantitative structure activity/property relationships (QSAR/QSPR) [13][14][15][16] for bitter taste [17,18] and antioxidant.…”