Definition of network types - Prediction of dough mechanical behaviour under shear by gluten microstructure

This study defines network types of wheat gluten to describe spatial arrangements of gluten networks in relation to dough mechanical behaviour. To achieve a high variety in gluten arrangements, ten specific and unspecific gluten-modifying agents in increasing concentrations were added to wheat dough. Gluten microstructure was visualized by confocal laser scanning microscopy and quantified by protein network analysis. Dough rheological behaviour was determined by both oscillatory and creep-recovery tests. Based on correlation matrices and principal component analysis, six different network types were identified and associated to their rheological characteristics: a cleaved (low viscous), rigid (highly viscous), spread (viscoelastic), strengthened (viscoelastic), particulate and dense (highly viscous) or particulate and loose (low viscous) network. Furthermore, rheological dough properties of specifically gluten-modified samples were predicted with five microstructural gluten attributes (lacunarity, branching rate, end-point rate, protein width, average protein length) and assigned properly by the obtained partial least square model with an accuracy up to 90% (e.g., R²Y = 0.84 for G*, 0.85 for tanδ, 0.90 for J_max). As a result, rheological properties of wheat doughs were predicted from microstructural investigations. This novel, quantitative definition of the relation between structure and mechanical behaviour can be used for developments of new wheat products with targeted properties.