A Machine Learning Approach to Predict Functional Performance From Measurable Protein Structural Characteristics: A Screening Tool for Protein Ingredient Quality

The food industry is witnessing the emergence of specialized protein-based functional ingredients for the use as gelling, thickening, and/or emulsifying agents in various food applications. Different sources of protein including species and cultivars, as well as variable processing conditions affect the protein's structural characteristics, which in turn govern their functional properties. The complex relationship between the structure and function of the protein can be modeled using machine learning (ML) algorithms. In this study, different ML algorithms were used to predict solubility, emulsifying activity index, emulsifying capacity, and gel strength of different plant proteins using structural predictors (surface hydrophobicity, zeta potential, undenatured protein content, water holding capacity, soluble protein polymer content, β-sheet content). Model performances were assessed by specific metrics ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow><mml:msup><mml:mi>R</mml:mi> <mml:mn>2</mml:mn></mml:msup> </mml:mrow> <mml:annotation>$$ {R}^2 $$</mml:annotation></mml:semantics> </mml:math> , mean absolute error [ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mi>MAE</mml:mi></mml:mrow> <mml:annotation>$$ \mathrm{MAE} $$</mml:annotation></mml:semantics> </mml:math> ], and root mean squared error [ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mtext>RMSE</mml:mtext></mml:mrow> <mml:annotation>$$ \mathrm{RMSE} $$</mml:annotation></mml:semantics> </mml:math> ]) and non-violation of physical constraints. The solubility and emulsifying activity index were predicted using surface hydrophobicity, zeta potential, and undenatured protein content. Emulsifying capacity was predicted using surface hydrophobicity, solubility, undenatured protein content, while gel strength was predicted using solubility, undenatured protein content, water holding capacity, soluble protein polymer content, and β-sheet content. The <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mtext>Gaussian</mml:mtext></mml:mrow> <mml:annotation>$$ Gaussian $$</mml:annotation></mml:semantics> </mml:math> based Support Vector Regression model accurately predicted solubility ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow><mml:msup><mml:mi>R</mml:mi> <mml:mn>2</mml:mn></mml:msup> </mml:mrow> <mml:annotation>$$ {R}^2 $$</mml:annotation></mml:semantics> </mml:math> = 0.8906), emulsifying activity index ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow><mml:msup><mml:mi>R</mml:mi> <mml:mn>2</mml:mn></mml:msup> </mml:mrow> <mml:annotation>$$ {R}^2 $$</mml:annotation></mml:semantics> </mml:math> = 0.7383), emulsifying capacity ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow><mml:msup><mml:mi>R</mml:mi> <mml:mn>2</mml:mn></mml:msup> </mml:mrow> <mml:annotation>$$ {R}^2 $$</mml:annotation></mml:semantics> </mml:math> = 0.7978), and gel strength ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow><mml:msup><mml:mi>R</mml:mi> <mml:mn>2</mml:mn></mml:msup> </mml:mrow> <mml:annotation>$$ {R}^2 $$</mml:annotation></mml:semantics> </mml:math> = 0.8822). Results highlighted the potential of ML algorithms for predicting of plant protein functionality using a few macromolecular structural characteristics. Such predictive models could serve as indispensable tools in the selection of protein ingredients for various food applications.