Rapid Determination Main Nutritional Components Of Coix Seeds Using Near Infrared Spectroscopy Technique And Study On Their Physicochemical Properties

Coix seed is the mature kernel of Coix lachryma-jobi L.,which is an annual or perennial herbaceous plant in the family of Gramineae.It contains many nutritional ingredients and functional components.The nutritive values of different varieties and region Coix seed（small Coix seeds（SCS）,big Coix seeds（BCS）and translucent Coix seeds（TCS））are diversity,now there is not rapid and effective screening method;The edible quality of Coix seed is influenced by the structure and properties of starch,and the previous studies lack specific state of the Coix seed varieties and comprehensive comparison for the different varieties;Prolamins are the major storage proteins in Coix seed and their genes have higher homology with zeins,but there are few reports about the structure and application quality of Coix seed prolamins.Therefore,this paper carried out distinguishing the different varieties Coix seed,determining nutritional ingredient by near infrared spectroscopy（NIRS）,and studying the structure and physicochemical properties of different varieties Coix seed starches and prolamins.The concrete content contents are as follows:（1）Rapid identification of Coix seed varieties and determination of starch content in Coix seed by NIRS and study on their physicochemical properties（1）Rapid identification of Coix seed varieties and determination of starch content in Coix seed by NIRSPrincipal component analysis（PCA）,learning vector quantization（LVQ）neural network and support vector machine（SVM）were used to carry out qualitative discriminant analysis.The scores plot of their principal components seriously overlapped and the categories of coix seed were difficult to distinguish.While the satisfactory results were obtained by LVQ neural network and SVM.The accuracy of LVQ neural network prediction is 90.91%,while the classification accuracy of SVM,whose penalty parameter and kernel function parameter were optimized,can be up to100%.Partial least squares regression（PLSR）was applied to establish quantitative models.Competitive adaptive reweighted sampling（CARS）and genetic algorithms（GA）were used as variable selection methods.Both of the two methods could decrease variables to a large extent,but the CARS-PLSR model obtained the best performances,the correlation coefficient（r）and residual predictive deviation（RPD）values were 0.94 and 2.41.It showed that the model could be used to make rough quantitative prediction.（2）Structure and physicochemical properties of different varieties Coix seed starchesThe structure and physicochemical properties of starch samples were considerable variation,and mainly attribute to amylose content.The BCS and SCS starches had lower amylose content,microporous structure,significantly higher relative crystallinity,water binding capacity,swelling power,solubility,thermal characteristics values,peak viscosity and breakdown than those of TCS starch.The all starch samples exhibited non-Newtonian shear-thinning flow behavior,represented a weak gel structure and had thixotropic loop.The TCS paste had the better viscoelasticity and the biggest hysteresis area;the SCS paste showed the best viscidity and lower hysteresis area;the BCS starch had the smallest viscoelasticity hysteresis area,which showed the best cold paste stability.（2）Rapid determination of fat content in Coix seed by NIRS and optimization of extraction conditions fatty acid from Coix seedNorris derivative smoothing（NDS）was used as pretreatment method,CARS was used as variable selection method and PLSR was applied to establish quantitative model for crude fat from different varieties and regions Coix seeds.The NDS and CARS removed the interference and useless information of original spectra.The R²,RMSECV,r,RMSEP and RPD values of the optimal PLSR model were 0.93,0.37,0.94,0.28 and 1.61,respectively.The PLSR model had the better robustness and this model could discriminate between low and high values of the response variable.The PLSR and backpropagation neural network（BPNN）were applied to establish models reflecting the relationship between fatty acid content and extraction conditions（temperature,time and extraction solvent）.Genetic algorithms（GA）and particle swarm optimization（PSO）were utilized to optimize the combination of extraction conditions.The fatty acid composition was analyzed by gas chromatography-mass spectrometer.The PLSR models well reflected the relation of FA content of BCS and SCS and extraction condition,while the BPNN model was more suitable for the TCS.The optimal extraction conditions of BCS（123℃,18min and acetone extraction）and SCS（126℃,20min and acetone extraction）were obtained by GA,and that of TCS（124℃,20min and acetone extraction）was gotten by PSO algorithms.The FA content of the three Coix seeds had slightly differences and the compostions were similar,namely,higher unsaturated fatty acid content.（3）Rapid determination of protein and amino acid content in Coix seed using NIRS technique and study on structure characteristics and properties of their prolamins（1）Rapid determination of protein and amino acid content in Coix seed using near infrared spectroscopy techniquePLSR was applied to establish quantitative model.CARS and NDS were used to improve the model accuracy.For protein,Asx,threonine,serine,Glx,alanine,leucine,proline,lysine,and histidine,NDS pretreatment improved the models’performance and yielded better prediction results.Then,key variables were selected by CARS,and the PLS models could obtain the optimal results with good predictive ability and robustness.The models of protein,Asx,serine,Glx,alanine,valine,isoleucine,leucine,phenylalanine,and proline were fit for screening（RPD≥2.5）,the threonine model（RPD=2.0）was fit for making rough quantitative prediction and the glycine,cysteine,methionine,tyrosine,arginine,lysine,and histidine models（RPD<1.5）might give a rough estimate of the contents of response variable.（2）Structure characteristics of different varieties Coix seeds prolamins and properties of their filmsThe three Coix seed prolamins showed the different protein content（TCS>SCS>BCS）,amino acid compositions and hydrophobicity（SCS>BCS>TCS）;their peptide sequences differences and the homology of Zea mays prolamins;the three prolamins had stable secondary structure,and the three prolamins aggregates were rodlike structures,formed very a fine meshwork and had self-assembly properties,which was the basis of their film-forming.The SCS film displayed better water absorption and higher elongation than those of the TCS film,while its film also showed the poor opacity and low glass transition temperature ascribed to the strong interaction of protein-protein and protein-water.The BCS prolamins could be denatured in the process of film casting,and its films had the worst properties.（4）Rapid determination of minerals and total dietary fiber content in Coix seed by NIRS and study on physicochemical properties of new dietary fibers（1）Rapid determination of minerals and total dietary fiber content in Coix seeds by near infrared spectroscopy technologyThe PLSR was applied to establish quantitative models for total dietary fiber（TDF）and mineral elements（mainly K,Mg,P and S）in Coix seed.The NDS was used as pretreatment method.The CARS,GA and random frog（RF）were used as variable selection methods.The results showed that CARS obtained the best performances of PLSR models.For the quantitative determination models of TDF,K,Mg,P and S,the optimal RPD values were 2.68,2.05,1.70,3.24 and 3.04,respectively.The PLSR models of TDF,P and S might be used to screen,the model of K be used to make rough quantitative prediction and the model of Mg only can discriminate between low and high values of the response variable.（2）Structure and physicochemical properties of new type of dietary fibersCoix seed starches underwent heat-moisture treatment（HMT）were new type of dietary fibers.Their structure and physicochemical properties were investigated.The amylose contents of BCS,TCS and SCS starches were 0.63%,24.12%and 0.96%,respectively.After HMT,starches appeared cracks and broken,birefringence weakened and the relative crystallinity and the ratio of 1047 cm^-1/1022 cm^-1 of HMT starches decreased,owing to disorganization of double helices and disruption of crystalline regions.The gelatinization temperatures of HMT starches increased but the thermal enthalpies decreased.The rheological characteristics had been changed,and HMT starches had the better flowability,especially for TCS starch,whose aging phenomenon was improved in some extent.HMT increased rapidly digestible starch and resistant starch（RS）levels.The TCS-HMT starch had the highest RS content.