Differential analysis method for quality characteristics of different varieties of millet
Through principal component analysis and cluster analysis methods, the quality indicators of Luliangshan Xiaomi were scored and classified, which solved the problem that the difference in quality characteristics of Xiaomi varieties was not effectively explored, and the quality evaluation and high-quality variety screening of Xiaomi varieties were realized. The high-value development provides a scientific basis.
Patent Information
- Application Number
- CN202410939929.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-15
- Publication Date
- 2024-11-08
AI Technical Summary
The existing technology has failed to effectively explore and analyze the quality characteristics and correlations of millet varieties in Luliangshan area, which has affected the quality evaluation of millet varieties and the screening of high-quality varieties, thereby limiting the high-value development and utilization of millet.
The principal component analysis and cluster analysis method were used to determine and analyze 15 quality indicators such as protein, fat, amylose, flavonoids, and yellow pigments of Luliangshan millet. The Pearson correlation coefficient, intergroup connection method, and Euclidean distance were used. Relevance analysis and data processing were carried out by other methods, key indicators were selected and comprehensive evaluation was conducted.
The differences in the quality characteristics of different varieties of Xiaomi in different regions have been successfully studied, providing millet processing adaptability, providing theoretical basis and scientific basis for the quality evaluation of millet varieties and screening of high-quality varieties, and promoting the high-value development and utilization of Xiaomi.
Smart Images

Figure CN118914484A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of biology and food, and particularly relates to a method for analyzing the difference of quality characteristics of different varieties of millet. Background Art
[0002] Millet is produced in many places, has a high yield, is cheap, and is rich in nutrients. Shanxi millet is known as the "crown jewel". Many studies have shown that millet has antioxidant function and contains a variety of active ingredients with health benefits such as lowering the three highs and anti-cancer, especially high content of polyphenols, flavonoids, and yellow pigments. It can be seen that millet plays an important role in optimizing consumers' dietary structure and meeting the diverse nutritional needs of the human body.
[0003] Although scholars have made a series of reports on the analysis and evaluation of various quality indicators of millet from different origins, millet varieties and planting areas have a great influence on various nutrients and functional components in millet. There are few reports on the analysis and evaluation of the quality of millet varieties in the Luliang Mountain area. Therefore, this paper explores the differences and correlations of millet quality indicators in the Luliang Mountain area, uses principal component analysis and cluster analysis methods to screen out key indicators that can be used to evaluate the quality of different millet varieties, and establishes an evaluation system, in order to provide a theoretical basis and scientific basis for the quality evaluation of millet varieties and the screening, introduction and breeding of high-quality varieties, which is of great significance to the high-value development and utilization of millet.
[0004] The Food and Agriculture Organization of the United Nations (FAO) declared 2023 as the "International Year of Millet", aiming to emphasize the further understanding of the nutritional and health value of millet, tap the huge potential of millet as a nutritious food, and the adaptability of millet under harsh and changing climatic conditions, encourage large-scale planting of millet to cope with climate change, and promote sustainable planting and consumption. Millet has the potential to become a new staple crop and have a place in people's dietary structure. In recent years, my country has proposed to establish a "big food concept", and the research and comprehensive evaluation of millet quality is also an important part of food security. To alleviate this problem, the present invention aims to comprehensively score Luliangshan millet through principal component analysis, use cluster analysis to verify the results of principal component analysis and classify millet varieties, in order to provide production guidance for market promotion and post-harvest processing. Summary of the invention
[0005] The purpose of the invention is to provide a method for analyzing the difference of quality characteristics of different varieties of millet, so as to study the difference of quality characteristics of different varieties of millet in different regions, thereby obtaining the processing adaptability of millet.
[0006] To achieve the above purpose, the present invention provides a method for analyzing the difference in quality characteristics of different varieties of millet, measures 15 quality indicators such as protein, fat, amylose, flavonoids, and yellow pigment in different millet, and evaluates the quality of Lvliang Mountain millet using principal component analysis and cluster analysis. Specifically, it includes the following steps:
[0007] S1. Collect different varieties of millet;
[0008] S2, pretreatment: the millet is sampled by quartering method, crushed by a grinder and passed through an 80-mesh sieve for later use;
[0009] S3. Determine the quality characteristics of millet, including protein, fat, amylose, vitamins, essential amino acids, etc.; normalize the data to ensure the accuracy and comparability of the data, and use coefficient of variation analysis to prove the stability of the data;
[0010] S4. Conduct difference analysis: including appearance quality of color and shape; including differences in nutritional components such as protein, amylose, vitamins, and essential amino acids;
[0011] Furthermore, the vitamins include vitamin A and vitamin B2; and the essential amino acids measured include methionine, valine, lysine, isoleucine and phenylalanine.
[0012] S5. Correlation analysis
[0013] Pearson correlation coefficient, intergroup linkage method and Euclidean distance were used to conduct correlation analysis on 15 quality indicators and essential amino acids of millet.
[0014] S6. Principal component analysis of quality characteristics of different millet varieties
[0015] After correlation analysis of the data, it was found that the data were significantly different and correlated, and there was information overlap, so principal component analysis (PCA) was performed on them. The collected data was standardized, and then PCA was used to extract the main components to reduce the data dimension and extract key indicators that can reflect the comprehensive quality of millet.
[0016] S7. Cluster analysis
[0017] Cluster analysis is to cluster the differences of different varieties and different indicators, so that those with similar distances are clustered into one category, so as to evaluate their comprehensive quality. Based on the comprehensive evaluation score of principal component analysis and the ranking of different millet varieties, the 24 millet varieties were classified using the inter-group connection method and Euclidean distance, and the results were compared with the principal component analysis scores to verify the accuracy of the principal component scores.
[0018] S8. Data processing
[0019] All indicators were measured three times, and SPSS software was used to analyze the data and draw PCA scree plots and loading plots. Origin2022 was used to draw PCA score plots, correlation heat maps, and cluster analysis heat maps. The online platform CNSknowall was used to draw polar coordinate stacked bar charts.
[0020] The invention has the following beneficial effects: the differences in quality characteristics of millet of different varieties in different regions are studied, thereby obtaining the processing adaptability of millet. The invention provides a theoretical basis and scientific basis for the quality evaluation of millet varieties and the screening, introduction and breeding of high-quality varieties, which is of great significance to the high-value development and utilization of millet. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 This is a polar coordinate stacked bar chart showing the differences in nutritional components of different millet varieties.
[0022] Figure 2 This is a heat map of the correlation between quality indicators of different varieties of millet.
[0023] Figure 3 Heat map showing the correlation between essential amino acids in different millet varieties.
[0024] Figure 4 This is the PCA scree plot.
[0025] Figure 5 This is the PCA loading diagram.
[0026] Figure 6 This is the principal component score diagram of different millet varieties.
[0027] Figure 7 Heat map for cluster analysis of different millet varieties. DETAILED DESCRIPTION
[0028] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments:
[0029] Example 1
[0030] S1. Collect different varieties of millet
[0031] Millet from 13 counties of Lvliang provided by the Economic Crops Research Institute of Shanxi Academy of Agricultural Sciences was used as the main raw material. The collection locations and varieties are shown in Table 2-1.
[0032] Table 2-1 Millet variety collection information
[0033]
[0034] Test reagents
[0035] Anhydrous ether, iodine solution, 200g·L-1 sodium hydroxide solution, alkaline copper tartrate solution, 2g·L-1 methyl red indicator, 85% ethanol, acetic acid solution, 0.02mol·L-1 potassium permanganate standard solution, anhydrous sodium sulfate, phenolphthalein indicator, chloroform, 10% amylase, 3% hydrogen peroxide solution, petroleum ether (boiling range 30-60°C), etc.
[0036] S2. Pretreatment: The millet was sampled by the quartering method, crushed with a grinder and passed through an 80-mesh sieve for the determination of millet quality characteristics;
[0037] S3. Determine the quality characteristics of millet, including protein, fat, amylose, vitamins, essential amino acids, etc.; normalize the data to ensure the accuracy and comparability of the data, and use coefficient of variation analysis to prove the stability of the data;
[0038] Determine the quality characteristics of different millet varieties in the Luliang Mountains, including protein, fat, grain size, color, etc.
[0039] (1) Appearance quality measurement
[0040] Determine the particle size and color (L* value, a* value, b* value) of millet.
[0041] (2) Nutritional composition determination
[0042] Protein: GB 5009.5 Determination of protein in food; Fat: GB 5009.6 Determination of fat in food; Amylose: GB / T 15683 Determination of amylose content in rice; Essential amino acids: GB 5009.124 Determination of amino acids in food; Minerals: GB 5009.92 Determination of calcium in food; GB 5009.87 Determination of phosphorus in food; Vitamins: GB5009.82 Determination of vitamins A, D and E in food; GB 5009.85 Determination of vitamin B2 in food;
[0043] S4. Conduct difference analysis: including appearance quality of color and shape; including differences in nutritional components such as protein, amylose, vitamins, and essential amino acids;
[0044] Furthermore, the vitamins include vitamin A and vitamin B2; and the essential amino acids measured include methionine, valine, lysine, isoleucine and phenylalanine.
[0045] 4.1 Differences in appearance and quality of different millet varieties
[0046] The appearance quality of millet is the most intuitive purchasing criterion for consumers when purchasing millet, and its appearance quality is mainly reflected in color, shape, etc. As shown in Table 3-1, the average particle size of different varieties of millet is distributed in the range of 1.16-1.72 mm, among which 2023 J32, 2023 B1, 2023 B14, and 2023 B22 are significantly different from other millet tested (P<0.05), but the changes are small. The color of millet is expressed by the lightness index, red-green value, and yellow-blue value (L*, a*, and b*), and the ranges are 52.63-63.49, 6.45-8.45, and 23.54-31.30, respectively. Among them, the color of 2023 variety J23 is the brightest, followed by 2023 variety J32, and there are significant differences in brightness between the two and other millet varieties; the brightness of 2023 variety J22 is significantly different from other millet varieties, and its L* value and b* value are the lowest among the tested millet varieties, and its b* value is also significantly different from other millet varieties, indicating that its color is the darkest and tends to be yellowish brown. The a* values of 2023 variety B1 and 2023 variety B22 are significantly different from other millet varieties (P<0.05), among which 2023 variety B1 has the highest a* value, tending to be reddish yellow. The coefficients of variation of particle size, L*, a*, and b* values were 8.89%, 3.63%, 6.14%, and 5.40%, respectively, indicating that the appearance of different millet varieties differed little.
[0047] Table 3-1 Differences in Xiaomi's appearance and quality
[0048]
[0049] Note: Different letters indicate significant differences (P<0.05)
[0050] 4.2 Differences in nutritional indicators of different millets
[0051] Table 3-2 Differences in nutritional content of different millets
[0052]
[0053] Combination Figure 1 As shown in Table 3-2, Jingu No. 21 has the highest protein content, followed by Jinfen No. 107, and the protein content of the two varieties is significantly different from that of other millet varieties (P<0.05). Among the 24 varieties, No. 2023 has the lowest protein content than No. 22. The protein content of different millet varieties varies from 9.32 to 11.86 g / 100 g. -1 The coefficient of variation is 7.77%, indicating that the degree of dispersion between the data is low and the differences between different millet varieties are small. 2023 J4, 2023 J8, 2023 J10, and Jingu No. 21 have higher fat content, among which Jingu No. 21 has the highest average content. Compared with other millets, the difference is more significant, and the fat content ranges from 3.14 to 4.84 g / 100g.-1 The coefficient of variation was 11.92%, indicating that the fat content of different millets was relatively small. The highest content of amylose in millet was 23.82 g / 100 g. -1 , minimum is 16.33g·100g -1 There are generally significant differences between varieties. The range of variation of vitamin A and vitamin B2 is 0.07-0.19 mg / 100g respectively. -1 , 0.01~0.05mg·100g -1 , the contents were generally low, with coefficients of variation of 29.15% and 41.54% respectively, indicating that there were large differences in vitamin content among millet species.
[0054] 4.3 Differences in polyphenols and mineral elements among different millet varieties
[0055] As shown in Table 3-3, the flavonoid content in millet is 8.42-11.29 mg·g -1 , among which 2023 variety J10 had the highest content, followed by 2023 variety J4 and 2023 variety J45, and 2023 variety J86 had the least content. The coefficients of variation of polyphenols and yellow pigments were high, 31.11% and 41.36% respectively, with large differences among species. Among them, 2023 variety J4, 2023 variety J23, 2023 variety J33, and Jingu 21 had higher polyphenol content and were significantly different from other varieties of millet. 2023 variety J4, 2023 variety J23, 2023 variety Bi1, 2023 variety Bi22, and Jingu 21 had more yellow pigment than other varieties. The yellow pigment content of 2023 variety J81 had significant differences among species (P<0.05) and was relatively low. According to the "High-quality Millet" published by the China Crop Science Association, the yellow pigment content of '2023 Pinbi 1', 'Jingu No. 21', and '2023 Pinbi 22' has reached the standard for high-quality millet. The range of mineral elements calcium and phosphorus is 152.5-218.6 mg kg -1 , 2146~3381mg·kg -1 Among them, the one with the highest calcium content is 2023 grade J45, and the one with the highest phosphorus content is 2023 grade B1.
[0056] Table 3-3 Differences in polyphenols and mineral elements in different millet varieties
[0057]
[0058] 4.4 Differences in essential amino acids among different millet varieties
[0059] It can be seen from Table 3-4 that the millet with the highest essential amino acids (EAA) is Jingu No. 21, followed by 2023 J4 and 2023 J82, and the lowest is 2023 Bi 22. The contents of methionine, valine, and isoleucine in Jingu No. 21 are the highest, and methionine and valine are significantly higher than those in other millet varieties, while the content of lysine is the lowest; the contents of methionine, phenylalanine, and leucine in 2023 J45 are the lowest, but the content of lysine is the highest; the contents of valine, isoleucine, phenylalanine, and leucine in 2023 J23 are the lowest, and the content of histidine is the highest; the contents of phenylalanine, leucine, and threonine in 2023 J10 are the highest; the content of tryptophan in 2023 J33 is significantly higher than that in other varieties (P<0.05). The coefficient of variation of essential amino acids is less than 20%, indicating that they are more concentrated in different millet varieties.
[0060] Table 3-4 Differences in EAA content in different millet varieties
[0061]
[0062]
[0063] S5. Correlation analysis
[0064] Pearson correlation coefficient, intergroup linkage method and Euclidean distance were used to conduct correlation analysis on 15 quality indicators and essential amino acids of millet. Figure 2 It can be seen that there is no significant correlation between the grain size and nutritional components of millet (P>0.05); protein is strongly correlated with the five quality indicators. Among them, it is positively correlated with fat and essential amino acids, and negatively correlated with amylose, Ca, and vitamin B2. This shows that there is a certain synergistic effect between protein and essential amino acids. Fat is negatively correlated with the three quality indicators. Among them, the correlation with amylose and Ca is the most significant, and the correlation with vitamin B2 is strong. Studies have shown that under certain conditions, fat and amylose can form a complex and affect the texture characteristics of millet, which may be closely related to the correlation between fat and amylose. Yellow pigment is significantly positively correlated with polyphenols, and significantly negatively correlated with vitamin A. By Figure 3 It can be seen that there is a generally highly significant positive correlation between essential amino acids, indicating that there is a certain synergistic effect between essential amino acids.
[0065] S6. Principal component analysis of quality characteristics of different millet varieties
[0066] Because the correlation between data will cause information overlap and thus lead to deviation in the results, after the data is standardized, it is necessary to perform PCA dimensionality reduction on the data. Extract multiple indicators into several key indicators for analysis to improve the accuracy of the results. From the correlation analysis, it can be seen that the quality characteristics of millet are not related to the size of millet particles, that is, the size of millet particles is not affected by the metabolic activity of internal nutrients. Therefore, when performing PCA, the millet particle size can be excluded to reduce interference. As shown in Table 3-5, from the initial eigenvalues and variance contribution rates of PCA, it can be seen that these five principal components integrate most of the information of the 14 indicators of millet quality.
[0067] Table 3-5 PCA initial eigenvalues and contribution rates
[0068]
[0069] Combination Figure 4 , Figure 5 , the initial eigenvalues and contribution rates of the principal components show that the variance contribution rate of the first principal component is 35.871%. Among the representative indicators that are positively loaded with the first principal component, essential amino acids have the largest contribution rate, with a load value of 0.972, followed by protein and fat, with load values of 0.971 and 0.936, respectively. Amylose and Ca are negatively loaded with the first principal component, with values of -0.964 and -0.906, respectively. Therefore, varieties with high essential amino acids and protein content and low amylose and Ca content perform better in the first principal component; the variance contribution rate of the second principal component is 19.339%, yellow pigment and a * The positive loading of the values is the largest, which are 0.926 and 0.928 respectively, while the vitamin A content shows a negative effect on the second principal component, with a loading value of -0.53, indicating that the key indicators synthesized by the second principal component are yellow pigment, a * value, vitamin A content. The variance contribution rate of the third principal component is 12.775%, L * Value, b * The values have a positive impact on the third principal component, with loading values of 0.648 and 0.605 respectively; while the P content has a negative impact of -0.661, indicating that L * Value, b * The value and P content have a higher weight in this principal component. The variance contribution rates of the fourth and fifth principal components are 7.791% and 7.201% respectively. Vitamin B2 has a positive effect on principal component 4, and flavonoids contribute the most to principal component 5, indicating that vitamin B2 and flavonoids are also important indicators for comprehensive evaluation of millet quality.
[0070] The cumulative contribution rate of the first and second principal components reached 55.21%, reflecting the most important indicator of millet quality characteristics. Therefore, the representative indicators of the first and second principal components are the core indicators affecting millet quality. From the loading diagrams of the first and second principal components and the results of principal component analysis, it can be concluded that the main indicators affecting millet quality are protein, fat, amylose, Ca, EAA, yellow pigment, and a* value.
[0071] Table 3-6 Principal component loading matrix
[0072]
[0073] Comprehensive evaluation of the quality of different millet varieties
[0074] Combination Figure 6 , a scoring model was constructed for the five principal components based on the loading matrix (Table 3-6). F1~F5 represent the scores of the principal components of different millet varieties. X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X12, X13, and X14 represent 14 quality indicators respectively, and F represents the comprehensive score of different varieties.
[0075] F1=0.433X1+0.418X2-0.430X3+0.119X4-0.034X5+0.029X6+0.044X7-
[0076] 0.229X8-0.404X9-0.079X 10 -0.124X 11 -0.047X 12 -0.082X 13 +0.434X 14
[0077] F2=0.029X1+0.029X2-0.063X3-0.122X4+0.358X5+0.563X6-0.322X7+
[0078] 0.002X8-0.033X9-0.027X 10 +0.053X 11 +0.564X 12 +0.326X 13 +0.041X 14
[0079] F3=0.031X1-0.057X2-0.093X3-0.268X4-0.177X5-0.034X6+0.414X7+
[0080] 0.155X8-0.108X9-0.494X10 +0.484X 11 -0.024X 12 +0.452X 13 +0.043X 14
[0081] F4=-0.041X1-0.038X2+0.013X3-0.005X4+0.476X5+
[0082] 0.029X6+0.353X7+0.608X8-0.208X9-0.231X 10 -0.243X 11 +0.062X 12 -
[0083] 0.334X 13 +0.052X 14
[0084] F5=-0.074X1+0.110X2+0.118X3+0.758X4+0.316X5-
[0085] 0.015X6+0.153X7-0.276X8+0.175X9-0.344X 10 +0.044X 11 -0.023X 12 +
[0086] 0.173X 13 -0.115X 14
[0087] The weights of each variable in the principal component were calculated by the coefficients of each variable in the component matrix and the eigenvalues of the principal component, and the scores of each principal component were obtained. Finally, the variance contribution rate of the principal component was used to establish a comprehensive evaluation model and the result was: F = 0.359F1 + 0.193F2 + 0.128F3 + 0.078F4 + 0.072F5. The comprehensive score characterizes the quality of different millet varieties. According to the final F value, different millet varieties in Luliang Mountain were evaluated. The top 10 varieties were 'Jingu No. 21', '2023 Pin J4', '2023 Pin J82', 'Jinfen No. 107', '2023 Pin J10', '2023 Pin J32', '2023 Pin J8', '2023 Pin J7', '2023 Pin J3', and '2023 Pin J81', with excellent quality characteristics. The top five millet varieties scored close to 1, which are high-quality millets.
[0088] Table 3-7 Principal component scores and comprehensive scores of different millet varieties
[0089]
[0090] S7. Cluster analysis (CA)
[0091] like Figure 7 ,On the basis of obtaining the comprehensive evaluation scores of principal component analysis and ranking the different millet varieties, ,the 24 millet varieties were classified using the inter-group linkage method and ,Euclidean distance, and the results were compared with the principal component analysis ,scoring results to verify the accuracy of the principal component scores.
[0092] Cluster analysis is to cluster the differences of different varieties and different indicators, so that those with similar distances are clustered into one category, so as to evaluate their comprehensive quality. Cluster analysis was performed on 14 quality indicators of 24 varieties of Luliang millet, such as Figure 3-7 As shown, different varieties of millet can be divided into four categories.
[0093] Category I: 2023 product J22. This type of variety L * Value, b * The value content is low, which means the color is darker and not easy to be liked by consumers, but the content of essential amino acids, flavonoids, polyphenols and P is high. This type of millet can be processed into millet pulp and mixed with cow's milk, homogenized, inoculated and fermented to make millet nutritional milk suitable for the elderly, the weak, women and children, and complementary foods for infants and young children.
[0094] Category II: '2023 Product Bi14', '2023 Product J85', '2023 Product J80', '2023 Product J42'. This type of millet contains polyphenols, yellow pigments, * The value content is generally low, which means that this type of millet has a low yellowness value and a generally high vitamin A content, which helps to improve immunity, promote growth and development, and is suitable for processing into VA-rich children's health food.
[0095] Category III: '2023 Pinbi 22', '2023 Pinbi 1', '2023 Pin J23', '2023 Pin J33', '2023 Pin J45', '2023 Pin J16', '2023 Pin J14', '2023 Pin J86', '2023 Pin J7'. These 9 types of millet have high contents of amylose and vitamin B2, among which '2023 Pinbi 22', '2023 Pin J23', and '2023 Pin J45' have the highest amylose content. This type of millet flour can be compounded with other grains to make meal replacement powder with a low GI value (glycemic index), and it has the potential to develop nutritionally fortified biscuits.
[31] , which is beneficial to consumers' blood sugar management and weight management. In addition, since amylose has good gel stability, water holding capacity and cooking quality, this type of millet can also be used to produce staple foods such as millet whole flour steamed bread and noodles.
[0096] Category IV: 'Jingu 21', '2023 Pin J4', '2023 Pin J32', '2023 Pin J82', '2023 Pin J101', '2023 Pin J81', '2023 Pin J10', 'Jinfen 107', '2023 Pin J8', '2023 Pin J3'. This type of millet generally has high protein, fat, and EAA content, and the unsaturated fatty acid content in millet fat is as high as 85.54%
[32] , mainly linoleic acid and oleic acid. This type of millet can be used to make millet porridge oil, which can provide a variety of choices for people with difficulty swallowing.
[33] .
[0097] Combined with the comprehensive scores of the principal components, it can be seen that in Category IV, 'Jingu No. 21', '2023 Pin J4', '2023 Pin J32', '2023 Pin J82', '2023 Pin J81', '2023 Pin J10', 'Jinfen No. 107', '2023 Pin J8', and '2023 Pin J3' are ranked in the top 10, and 2023 Pin J101 ranks 12th, with the highest cumulative comprehensive score and the best quality, belonging to high-quality millet; In Class II, except for '2023 Pin J42' which ranked 11th, the other three millet varieties ranked 20th, 22nd, and 24th respectively; in Class III, except for '2023 Pin J7' which ranked 8th and '2023 Pin Bi 22' which ranked 13th, the other millet varieties ranked last. Class II and Class III belong to low-quality millet with the worst quality; Class I '2023 Pin J22' had a comprehensive score of F=-0.10, ranked 13th, and belonged to medium-quality millet with average quality. The results of principal component analysis were verified by cluster analysis, and the results were consistent.
[0098] S8. Data processing
[0099] All indicators were measured three times, and SPSS software was used to analyze the data and draw PCA scree plots and loading plots. Origin2022 was used to draw PCA score plots, correlation heat maps, and cluster analysis heat maps. The online platform CNSknowall was used to draw polar coordinate stacked bar charts.
[0100] This specific embodiment is merely an explanation of the present invention and is not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by the patent law.
Claims
1. A method for analyzing the difference in quality characteristics of different varieties of millet, characterized in that: Fifteen quality indicators including protein, fat, amylose, flavonoids, and yellow pigments were measured in different millets, and the quality of Lvliangshan millet was evaluated using principal component analysis and cluster analysis, which specifically included the following steps: S1. Collect different varieties of millet; S2, pretreatment: the millet is sampled by quartering method, crushed by a grinder and passed through an 80-mesh sieve for later use; S3. Determine the quality characteristics of millet, including protein, fat, amylose, vitamins, essential amino acids, etc.; normalize the data to ensure the accuracy and comparability of the data, and use coefficient of variation analysis to prove the stability of the data; S4. Conduct difference analysis: including appearance quality of color and shape; Including differences in nutritional content of protein, amylose, vitamins, and essential amino acids; S5. Correlation analysis: Pearson correlation coefficient, intergroup linkage method and Euclidean distance were used to conduct correlation analysis on 15 quality indicators and essential amino acids of millet; S6. Principal component analysis of millet quality characteristics of different varieties: After correlation analysis of the data, it was found that the data were significantly different and correlated, and there was information overlap. Therefore, principal component analysis (PCA) was performed on the data. The collected data was standardized, and then PCA was used to extract the main components, reduce the data dimension and extract key indicators reflecting the comprehensive quality of millet. S7. Cluster analysis: Cluster analysis is to cluster the differences of different varieties and different indicators so that those with similar distances are clustered into one category, so as to evaluate their comprehensive quality. On the basis of obtaining the comprehensive evaluation score of principal component analysis and ranking different millet varieties, the inter-group connection method and Euclidean distance were used to classify 24 millet varieties. The results were compared with the principal component analysis scores and the accuracy of the principal component scores was verified. S8. Data processing.
2. The method for analyzing the difference of quality characteristics of different varieties of millet according to claim 1, characterized in that: The vitamins in the S4 include vitamin A and vitamin B2; the essential amino acids measured include methionine, valine, lysine, isoleucine and phenylalanine.
3. The method for analyzing the difference of quality characteristics of different varieties of millet according to claim 1, characterized in that: The data processing method in the S8 is to measure all indicators three times, use SPSS software to analyze the data and draw PCA scree plots and loading plots, use Origin2022 to draw PCA score plots, correlation heat maps, and cluster analysis heat maps, and use the online platform CNSknowall to draw polar coordinate stacked bar charts.
