No abstract

Background and objectives The effect of defatting four oilseed protein concentrates (safflower, sunflower, canola, and hemp) on the surface and functional properties of the proteins was investigated as a function of pH (pH 3, 5, 7). The functionality of commercial protein concentrates (soy, faba bean, lentil, pea, northern great bean, whey) already in the marketplace was also tested for comparative purposes. Findings Defatting with hexane increased the protein content from 77.3% to 92.9% for safflower, 67.5 to 75.5% for sunflower, 58.0–66.0% for canola, and 71.0–83.2% for hemp. The approximate isoelectric point (pI) of safflower increased with defatting (5.4–5.8), whereas for canola the pI decreased with defatting (4.7–4.3), and sunflower and hemp protein concentrates had similar pI for defatted or full fat. Certain functional properties were improved with defatting, whereas others showed the opposite trend; this was highly dependent on protein type and pH. Conclusions The oilseed concentrates were comparable to the concentrates in the marketplace with canola and sunflower proteins having the greatest oil‐holding capacity and defatted safflower having the highest foaming capacity of all the proteins tested. Significance and novelty Based on their functionality, the oilseed protein concentrates have potential to be used by the food ingredient industry.

No abstract

No abstract

Machine learning methods such as multilayer perceptrons (MLP) and Convolutional Neural Networks (CNN) have emerged as promising methods for genomic prediction (GP). In this context, we assess the performance of MLP and CNN on regression and classification tasks in a case study with maize hybrids. The genomic information was provided to the MLP as a relationship matrix and to the CNN as “genomic images.” In the regression task, the machine learning models were compared along with GBLUP. Under the classification task, MLP and CNN were compared. In this case, the traits (plant height and grain yield) were discretized in such a way to create balanced (moderate selection intensity) and unbalanced (extreme selection intensity) datasets for further evaluations. An automatic hyperparameter search for MLP and CNN was performed, and the best models were reported. For both task types, several metrics were calculated under a validation scheme to assess the effect of the prediction method and other variables. Overall, MLP and CNN presented competitive results to GBLUP. Also, we bring new insights on automated machine learning for genomic prediction and its implications to plant breeding.