Pankaj Bhowmik

Background and Objectives: The goal of this research was to examine the relationship between the composition and functionality of pea flour using the following machine learning algorithms: linear regression, partial least squares regression (PLSR), Gaussian process regression (GPR), support vector regression, gradient‐boosted decision trees, and a standard feed‐forward neural network. Findings: In general, linear models outperformed non‐linear models. PLSR provided best fits for prediction of emulsion stability, oil holding capacity, foam stability and foam capacity; but was less effective for solubility and water holding capacity, which were best described by the GPR model. Variable Importance in Projection scores, calculated for each PLSR model, showed that protein and acid detergent fiber were both highly influential in predicting foaming capacity (1.52 and 1.55), foaming stability (1.30 and 1.54), oil holding capacity (1.64 and 1.50), and water holding capacity (1.56 and 1.53). Protein was also highly important in predicting solubility (1.80), alongside starch (1.60) whereas lipid was highly predictive (2.02) for emulsion stability. Conclusion: Application of machine learning models was successful in relating compositional features of pea flour to functionality. Significance and Novelty: Using machine learning to predict the functional behavior of pea will aid both breeders and product developers in ingredient selection.

Sixteen commercially wet‐extracted protein isolates and 13 commercially air‐classified protein concentrates were evaluated for their particle size, physicochemical properties (protein solubility, differential scanning calorimetry, surface hydrophobicity, and surface charge) and functional properties (water holding capacity, oil absorption capacity, emulsifying capacity and stability, foaming capacity and stability, and least gelling concentration). Protein concentrates exhibited lower average values for particle size, surface charge, surface hydrophobicity, and water holding capacity, but higher values for oil absorption capacity, protein solubility, emulsion capacity and stability, as well as foaming capacity and stability, compared to protein isolates. Intensive processing appeared to denature proteins in most isolates, leading to reduced functional properties. Changes in pH and the addition of 2% NaCl influenced the properties of both protein isolates and concentrates. Protein solubility, emulsion capacity and stability, and foaming stability of the protein samples were superior at pH 10, followed by pH 3. The addition of 2% NaCl enhanced foaming capacity and reduced the least gelling concentration. A significant interaction between sample and treatment conditions ( p < 0.001) in protein isolates and protein concentrates was observed. Crop type differences in protein solubility, oil emulsion capacity and foaming stability of protein concentrates under certain conditions were detected using maximum likelihood factor analysis (MFLA). However, these differences were not evident in protein isolates, possibly due to more variable processing conditions.

Yellow peas (YPs) are a promising source of sustainable plant proteins however, characteristic off‐flavors limit consumer acceptance and hinder their use as value‐added food ingredients. This study evaluated the compositional and sensory characteristics of fourteen processed YP ingredients to determine their suitability in pan breads. Nine protein isolates (YPIs), three protein concentrates (YPCs), and two heat‐treated flours (YP‐IR and YP‐RF) were analyzed for fatty acid composition, lipoxygenase (LOX) activity, volatile organic compounds (VOCs), and electronic nose (eNose) responses. LOX activity and linoleic acid, a key LOX substrate, differed significantly between YP ingredients. Fifteen VOCs previously linked to off‐flavors were identified, with YPIs exhibiting the highest concentrations. eNose analysis correctly predicted YP ingredient identities with 69.5% accuracy and distinguished treated flours from other YP ingredients. Pan breads using two YPI ingredients (YPI‐1 and YPI‐2) predicted to produce strong off‐flavors, the two heat‐treated flours and a wheat flour (WF) control were prepared. Breads were evaluated for proximate composition, consumer acceptance, descriptive sensory attributes, and instrumental color. YPI‐1 bread was significantly less acceptable overall than WF ( p < 0.001) and showed higher pea aroma and flavor ( p < 0.001). YPI‐2 bread had a denser texture ( p < 0.05) and significantly higher wheaty aroma ( p < 0.01). Breads made with YP‐IR and YP‐RF performed closest to the WF control. The heat‐treated YP flours demonstrated the greatest potential as pan bread additives. YPI samples were associated with undesirable sensory characteristics. Analyses of the raw YP ingredients effectively predicted sensory differences, highlighting the usefulness of these methods as screening tools during product development. Practical Applications: This research shows that different value‐added yellow pea ingredients can introduce a range of flavors into food matrices, and that flavoromics tools can assist with product development. The two heat‐treated yellow pea flours were identified as the best options for bread making because they were less related to pea aroma and flavor in the final product. These findings can help bakers and food manufacturers choose pea ingredients that make high‐protein breads taste better.

The advent of artificial intelligence (AI) holds great promise for revolutionizing the fields of plant tissue culture and genome editing. Plant tissue culture is recognized as a powerful tool for rapid multiplication and crop improvement. However, the complex interactions between genetic and environmental factors generate large volumes of data, posing challenges for traditional statistical analysis methods. To address this, researchers are now employing machine learning (ML)-based and artificial neural networks (ANN) approaches to predict and optimize in vitro culture protocols thereby improving precision, sustainability, and efficiency. Integrating AI technologies such as machine learning (ML), artificial neural networks (ANN), and deep learning (DL) can significantly advance the development of data-driven models for CRISPR/Cas9 genome editing. Today, AI-driven methods are routinely applied to enhance precision in predicting on- and off-target sequence locations and editing outcomes. Additionally, predicting protein structures can provide a directed evolution framework that facilitates the creation of improved gene editing tools. However, the application of AI-based CRISPR modeling in plants is not yet fully explored. In this context, we aim to examine representative ML/DL/ANN models of CRISPR/Cas based editing employed in various organisms. This review significantly compiles a diverse set of studies and provides a clear overview of how AI is transforming the fields of plant tissue culture and genome editing. It emphasizes AI’s potential to increase the efficiency and precision of biotechnological practices, making them more accessible and cost-effective. While outlining current findings, the paper sets the stage for future research, encouraging further exploration into the integration of AI with plant biotechnology.

Chloroplast transformation technology has become a powerful platform for generating plants that express foreign proteins of pharmaceutical and agricultural importance at high levels. Chloroplasts are often chosen as attractive targets for the introduction of new agronomic traits because they have their own genome and protein synthesis machinery. Certain valuable traits have been genetically engineered into plastid genomes to improve crop yield, nutritional quality, resistance to abiotic and biotic stresses, and the production of industrial enzymes and therapeutic proteins. Synthetic biology approaches aim to play an important role in expressing multiple genes through plastid engineering, without the risk of pleiotropic effects in transplastomic plants. Despite many promising laboratory-level successes, no transplastomic crop has been commercialized to date. This technology is mostly confined to model species in academic laboratories and needs to be expanded to other agronomically important crop species to capitalize on its significant commercial potential. However, in recent years, some transplastomic lines are progressing in field trials, offering hope that they will pass regulatory approval and enter the marketplace. This review provides a comprehensive summary of new and emerging technologies employed for plastid transformation and discusses key synthetic biology elements that are necessary for the construction of modern transformation vectors. It also focuses on various novel insights and challenges to overcome in chloroplast transformation.

The nutrient composition and in vitro digestibility of twenty‐seven commercial pulse protein isolates (PI) and protein concentrates (PC) derived from pea, faba, and lentil, and two soy protein isolate controls were tested using consistent analytical methods to understand compositional variability amongst products manufactured by different suppliers. Principal component analysis (PCA) and maximum likelihood factor analysis (MLFA) were applied to model the compositional and amino acid data to determine where the variability between protein products lay. MLFA delineated between isolates and concentrates based upon protein and total dietary fibre content, while moisture and fat variability could be used to differentiate samples within the PI or PC groupings. PCA score charts distinguished between isolates and concentrates due to higher relative concentrations of amino acids in the isolates, with glutamine/glutamic acid contributing to this distinction. Separation according to crop type within PI and PC groupings based upon the arginine and phenylalanine content was also evident. Sodium, potassium, magnesium, and iron micronutrients also contributed to the variability between PI and PC samples. Calculated amino acid scores showed all samples contained sufficient concentrations of essential amino acids to meet FAO requirements established for preschool‐aged children. PI samples had higher in vitro digestibility than PC samples with minimal variability within the groupings.