A method for broad-spectrum, linear, and accurate prediction of early peanut kernel sucrose content based on near-infrared spectroscopy

Abstract

This invention discloses a method for broadly linearly and accurately predicting the sucrose content of early-stage peanut kernels based on near-infrared spectroscopy. It employs near-infrared spectroscopy technology, constructs a data prediction model based on sucrose content gradient data from standard samples, and combines this with a linear optimal algorithm for data fitting. Furthermore, it utilizes peanut generation populations of kernels and optimizes the data model based on observable and comparable single or multiple quantitative indicators. This invention employs optimized data steps and chemical process parameters, and uses a modified independent orthogonal first and second-order differential parameter dual-decision data model. This effectively and linearly broadens the accurate prediction range of sucrose content in peanut kernels based on a small number of kernels. The resulting algorithm enables rapid and non-destructive determination of sucrose content in individual plants of early-generation hybrids, effectively improving the breeding efficiency of high-sucrose-content peanut varieties.

Description

Technical Field

[0001] This invention relates to the quality measurement of peanuts and their application in industry, agriculture and peanut breeding, and in particular to a systematic method that linearly expands the range of accurate prediction of sucrose content in peanut kernels for a small number of kernels. Background Technology

[0002] Peanuts, as an important economic crop, are widely cultivated in over 100 countries. Peanut kernels are rich in various vitamins, carbohydrates, proteins, essential fatty acids, minerals, and other nutrients. Especially today, with global climate change causing food security issues, edible peanuts, as a food crop that can meet human nutritional needs, are of great significance in maintaining food security. Sugars not only play an important role in human life processes but also have a significant impact on the quality and processed flavor of edible peanuts. Studies have shown that the sugar content of peanut kernels can serve as an important indicator for predicting the flavor and sweetness of roasted peanuts, and even small changes can affect the final roasting quality. Therefore, increasing the sugar content of peanut kernels is an important means of improving the quality of edible peanuts. Currently, traditional methods for determining sugar content are mostly colorimetric methods, such as the 3,5-dinitrosalicylic acid colorimetric method and the Fehling's reagent colorimetric method. Colorimetric methods often have poor specificity and low detection efficiency, and are not suitable for the simultaneous analysis of large numbers of samples. Li Weitao et al. established a method for simultaneously determining the fructose, glucose, and sucrose content in peanut kernels using high-performance liquid chromatography-refractive index (HPLC-RID). Analysis of 20 peanut varieties revealed that the main soluble sugar in peanut kernels was sucrose, with no detected fructose or glucose. Therefore, targeting sucrose content in peanut kernels as a key trait, and exploring and cultivating peanut germplasm and new varieties with high sucrose content, is of great significance for improving the quality of edible peanuts.

[0003] Existing research has shown that sucrose content in peanut kernels is a heritable quantitative trait. High-throughput, large-scale detection and analysis of sucrose content in resource materials and breeding progeny is a key technology for breeding new edible peanut varieties with high sucrose content. Near-infrared spectroscopy, with its non-destructive, rapid, and efficient characteristics, has been widely used in the analysis of important quality indicators in peanuts, such as oil content, fatty acids, protein, and amino acids.

[0004] However, existing technologies have limited sucrose content ranges among peanut materials during data modeling, ranging from 2.82% to 7.11% and 2.15% to 7.93%, respectively. In particular, the proportion of peanut materials with high sucrose content is small, which to some extent limits the accuracy of the models used in breeding. Furthermore, the number of peanut kernels required for constructing near-infrared models and predicting sucrose content is typically 30 to 50, making it impossible to detect sucrose content in individual plants with fewer kernels in early generations of breeding. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the various shortcomings of the prior art and provide a method for broadly linearly and accurately predicting the sucrose content of early peanut kernels based on near-infrared spectroscopy under the condition of a small amount of peanut kernels. This method is suitable for situations where there are few peanut kernels in the early stage of breeding, and in particular, it broadens the predictable range of sucrose content.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows.

[0007] This method, based on a few peanut kernels, linearly expands the accurate prediction range of sucrose content in peanut kernels. It is used for early-stage varietal trait screening and / or rapid, non-destructive identification of peanut kernel sweetness in agriculture / industry. The method employs near-infrared spectroscopy, constructing a data prediction model based on sucrose content gradient data from standard samples and a linear optimal algorithm for data fitting. Furthermore, it utilizes peanut kernels from different generations and optimizes the data model based on observable and comparable single or multiple quantitative indicators. This expands the accurate linear predictable range of sucrose content in peanut kernels under technical parameters where available peanut kernels are limited or normal.

[0008] As a preferred embodiment of the present invention, the data prediction model is constructed using partial least squares and / or modified least squares methods based on first-order and / or second-order differential infrared spectral data as the linear optimal algorithm for data fitting.

[0009] As a preferred technical solution of the present invention, the improved least squares method is to establish a mathematical model based on the chemical value of sucrose content and the near-infrared spectral data value of peanut kernels in the standard sample set using partial least squares (PLS) based on the first-order and second-order differential infrared spectral data respectively; the main model is constructed using the first-order differential and its derivative, and the auxiliary model is constructed using the second-order differential and its reciprocal; when fuzziness of quantitative indicators occurs under the main model in the subsequent model optimization data process, the auxiliary model is referred to for auxiliary decision-making, thereby improving the construction efficiency of the data model and improving the robust accuracy of the representative model's inherent reliable measurement.

[0010] As a preferred embodiment of the present invention, the optimization of the data prediction model is based on a quantifiable dual indicator, namely the coefficient of determination R. 2 The root mean square error (RMSEP) is optimized.

[0011] As a preferred embodiment of the present invention, the optimization of the data prediction model uses the final value of its dual indicators as the coefficient of determination R. 2＝ 0.9054±0.004, Root Mean Square Error (RMSEP) ＝ 0.6774±0.004; the corresponding data prediction model accurately predicts the sucrose content of peanut kernels in a linear range of ≤12% w / w, especially 7%-12% w / w.

[0012] As a preferred embodiment of the present invention, the minimum number of available peanut kernels is no more than 18-20.

[0013] As a preferred embodiment of the present invention, the method includes the following steps:

[0014] A. Take a peanut standard sample set and keep it at a constant temperature of about 25℃ for more than 48 hours. Then, use a near-infrared quality analyzer to perform spectral determination on the samples, scanning wavelength range of 950-1650nm and resolution of 5nm. Take 18-23 peanut kernels from each sample and place them evenly into a small sample cup with a diameter of 6cm, fully covering the bottom mirror surface. Repeat the sample loading 3 times, and measure twice for each loading. A total of 6 spectral data points are obtained for each sample. The average spectrum is used for model construction. Remove the red skin from the peanut kernels used for spectral collection, grind them with a grinder, and pass them through a 20-mesh sieve. Take 0.1g of powder sample, accurately weigh it, place it in a 2mL centrifuge tube, defatt it twice with n-hexane, and then mix it at a ratio of 1g:15mL. Add 50% v / v ethanol aqueous solution precisely at a w / v ratio, mix well, soak for 30 min, and extract with ultrasonic-assisted shaking at 250W-40kHz at room temperature for 30 min; centrifuge at 12000r / min at room temperature for 10 min; accurately measure 1 mL of supernatant into a 1.5 mL centrifuge tube, and accurately add 50 μL each of zinc acetate solution and potassium ferrocyanide solution. After the white zinc ferrocyanide precipitate is completely formed, centrifuge at 12000r / min for 20 min, collect the supernatant, and determine the sucrose content in peanut kernels using HPLC-RID method;

[0015] B. Existing chemometrics software was used to process the first derivative and standard normal transformation of the spectral information of the standard sample set. Partial least squares (PLS) was employed to establish a mathematical model for the sucrose content chemical values ​​and near-infrared spectral data of peanut kernels in the standard sample set. Internal cross-validation was repeatedly used to eliminate outliers. Simultaneously, parameters were modified to construct a second-order differential dataset orthogonal to the first-order derivative of the spectral data. Through the same data processing process, a new independent auxiliary data model was obtained. Orthogonality and independence ensured that the second-order auxiliary data model played a crucial role in assisting and supplementing the fuzzy decision-making position of the first-order master data model. This improved the efficiency of data model construction, especially enhancing the robust accuracy representing the model's inherent measurement reliability. This was particularly effective in extending the linear and accurate prediction range of sucrose content in peanut kernels when only a limited number of measurable peanut kernels were available. Finally, the coefficient of determination R of the model was cross-referenced. 2 The root mean square error (RMSEP) is used to optimize the data model; the final value is the coefficient of determination R. 2＝ 0.9054, Root Mean Square Error (RMSEP) ＝0.6774; The final data model obtained can be packaged and applied after external verification.

[0016] As a preferred technical solution of the present invention, in step A, the optional method for constructing the standard sample set is as follows: 325 families in the RIL population constructed by hybridizing the high-yielding and superior line 201138 and Jihuatian 1 are numbered and used as the standard sample set; peanut sample materials are naturally air-dried to a moisture content of less than 5%, and mature, plump, undamaged and unmolded peanut seeds are selected as the test samples.

[0017] In a preferred embodiment of the present invention, in step A, the sucrose content in peanut kernels is determined by HPLC-RID method. The chromatographic column is an Agilent Zorbax carbohydrate column, 4.6 mm × 250 mm, 5 μm. The mobile phase is acetonitrile-water, 70:30, v / v, the flow rate is 1.0 mL / min, the column temperature and the differential refractive index detector temperature are both 40 °C, and the injection volume is 10 μL. The sucrose content % = sucrose concentration (mg / mL) × 15 mL × 0.1. The sucrose concentration of the sample is calculated based on the peak area of ​​the sample and the sucrose standard curve. Each sample is measured in parallel three times, and the average value is taken.

[0018] As a preferred technical solution of the present invention, in step B, the optional method for external verification is as follows: randomly select 20 families from the RIL population of Jihua 11 and Jihuatian 1, numbered ST01 to ST20, use the constructed final value data model to predict the sucrose content in the kernel, and compare it with the sucrose content determined by HPLC-RID method to complete the external verification of the data model.

[0019] The beneficial effects of the above technical solution are as follows: This invention employs optimized data steps and chemical process parameters, based on a small sample cup containing 18-23 peanuts and partial least squares method, and uses improved independent orthogonal first and second-order differential parameters to construct a dual-decision data model. A second-order differential dataset orthogonal to the first derivative of the spectral data is constructed, and a new independent auxiliary data model is obtained through the same data processing process. The orthogonality and independence of the second-order auxiliary data model provide important auxiliary and supplementary data processing benefits at the fuzzy decision position of the first-order main data model. This improves the efficiency of data model construction, especially enhancing the robust accuracy representing the inherent reliability of the model's measurement. It also plays a substantial role in effectively expanding the linear and accurate prediction range of peanut sucrose content using the data model when only a few measurable peanut kernels are available. Overall, the near-infrared detection model for peanut sucrose content constructed in this invention has been externally validated, and its predicted values ​​and chemical values ​​have a certain coefficient of determination R. 2The value of 0.9478 indicates that the model accurately predicts the sucrose content of peanut kernels, enabling rapid and non-destructive determination of sucrose content in individual plants during the early generations of hybridization, and effectively improving the breeding efficiency of peanut varieties with high sucrose content. Attached Figure Description

[0020] Figure 1 The near-infrared instrument and sample cup used in this invention.

[0021] Figure 2 This is the near-infrared spectrum of peanut kernels.

[0022] Figure 3 Typical chromatograms of three solutions are shown; where: blank solution (A), reference solution (B), and test solution (C).

[0023] Figure 4 Scatter plot of near-infrared spectral prediction and chemical values ​​of sucrose content in peanut kernels.

[0024] Figure 5 This diagram illustrates the correlation between the predicted sucrose content and the chemical value of peanut kernels for external verification. Detailed Implementation

[0025] The following embodiments illustrate the present invention in detail. All raw materials and equipment used in the present invention are conventional commercially available products and can be directly obtained through market purchase. In the following description of the embodiments, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that the present application can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, and methods are omitted so as not to obscure the description of this application with unnecessary detail.

[0026] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of a described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. It should also be understood that, as used in this specification and the appended claims, the term "and / or" refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0027] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrases "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]." Furthermore, in the description of this specification and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0028] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0029] This invention uses a recombinant inbred line (RIL) population constructed by hybridizing the high-yielding and superior lines 201138 and Jihuatian 1 as materials. The sucrose content in the kernels of 325 families within the RIL population was determined using HPLC-RID. A near-infrared spectroscopy (NIRS) model for peanut sucrose content was constructed using a small sample cup (18-23 kernels) and partial least squares (PLS) and a modified double PLS method, targeting a DA7200 near-infrared quality analyzer from Boton GmbH, Sweden. The accuracy of the constructed model was verified by randomly selecting 20 families from the Jihuatian 11 and Jihuatian 1 RIL populations. The aim is to provide technical support for the rapid detection of sucrose content in peanut resources and for the efficient breeding of high-sucrose-content edible peanuts.

[0030] Example 1

[0031] A near-infrared model for peanut kernel sucrose content was constructed using 325 families from a RIL population (constructed through hybridization of the high-yielding and superior line 201138 and Jihuatian 1) as the standard sample set, numbered S001–S325. Twenty families from the Jihua 11 and Jihuatian 1 RIL populations (numbered ST01–ST20) were randomly selected for external model validation to assess the model's accuracy. All experimental materials were provided by the Peanut Research Laboratory of the Grain and Oil Crops Research Institute, Hebei Academy of Agricultural and Forestry Sciences.

[0032] Example 2

[0033] All peanut materials were naturally air-dried to a moisture content below 5%. Mature, plump, and undamaged peanut seeds were selected as test samples and kept at a constant temperature of approximately 25℃ for at least 48 hours. The samples were then subjected to spectral analysis using a DA7200 near-infrared quality analyzer (Swedish company, Böton), with a scanning wavelength range of 950–1650 nm and a resolution of 5 nm. After the instrument preheated, 18–23 kernels from each sample were placed evenly into a 6 cm diameter sample cup, ensuring the bottom mirror surface was fully covered. Figure 1 The sample was loaded three times, and the sample was measured twice each time. A total of six spectral data points were obtained for each sample, and the average spectrum was used for model construction.

[0034] See Figure 2 The near-infrared spectra of peanut kernels in the standard sample set collected in this study are as follows: Figure 2 As shown, in the range of 950–1650 nm, the near-infrared curves of each sample exhibit multiple absorption peaks. Although the trends are roughly the same, the peak values ​​of each sample show significant differences.

[0035] Example 3

[0036] Remove the red skin from the peanut kernels used for spectral acquisition, grind them using a grinder, and pass them through a 20-mesh sieve. Accurately weigh approximately 0.1 g of the powder sample and place it in a 2 mL centrifuge tube. Defatt the sample twice with n-hexane, then accurately add 50% ethanol aqueous solution (v / v) at a ratio of 1 g:15 mL (w / v). Mix well and soak for 30 min. Extract using ultrasonic-assisted extraction at room temperature (250 W, 40 kHz) for 30 min, shaking every 10 min during extraction. Centrifuge at 12000 rpm for 10 min at room temperature. Accurately measure 1 mL of the supernatant into a 1.5 mL centrifuge tube, and accurately add 50 μL each of zinc acetate solution and potassium ferrocyanide solution. Once the white zinc ferrocyanide precipitate is completely formed, centrifuge at 12000 rpm for 20 min. Use the supernatant for liquid chromatography analysis. Specifically, the sucrose content in peanut kernels was determined using HPLC-RID. An Agilent Zorbax carbohydrate column (4.6 mm × 250 mm, 5 μm) was used, with acetonitrile-water (70:30, v / v) as the mobile phase, a flow rate of 1.0 mL / min, and a column temperature and differential refractive index detector temperature of 40 °C. The injection volume was 10 μL. Sucrose content (%) = sucrose concentration (mg / mL) × 15 mL × 0.1. The sucrose concentration of the sample was calculated based on the peak area and the sucrose standard curve. Each sample was measured in triplicate, and the average value was used for model construction and validation.

[0037] See Figure 3 The experiment yielded typical chromatograms of the blank solution and the sucrose reference solution. Figure 3 A, B). The peanut kernels were prepared as a test solution according to the procedure in "1.2.2 Determination of Sucrose Content in Standard Sample Set". The solution was analyzed by liquid chromatography, and a typical chromatogram of the test solution was obtained. Figure 3 C). As can be seen, the sucrose content determination results of 325 peanut kernel samples from the standard sample set RIL population are shown in Table 1.

[0038] Table 1. Results of sucrose content determination in 325 peanut kernel samples from the standard sample set RIL population.

[0039]

[0040] As can be seen, the average sucrose content is 5.69%, the sucrose content ranges from 2.07% to 12.37%, and the coefficient of variation is 38.80%. The sucrose content has a wide distribution range and a large coefficient of variation, which meets the requirements for constructing a near-infrared model of sucrose content.

[0041] Example 4

[0042] Based on existing literature, this study employs existing chemometric software to perform first-order derivative processing and standard normal transformation on the spectral information of the standard sample set. Partial least squares (PLS) is used to establish a mathematical model for the sucrose content chemical values ​​and near-infrared spectral data of peanut kernels in the standard sample set. Internal cross-validation is repeatedly used to eliminate outliers. Simultaneously, parameters can be modified to construct a second-order differential dataset orthogonal to the first-order derivative of the spectral data. Through the same data processing process, a new independent auxiliary data model is obtained. The orthogonality and independence of the second-order auxiliary data model provide significant auxiliary and supplementary data processing benefits at the fuzzy decision position of the first-order main data model. This improves the efficiency of data model construction, especially enhancing the robust accuracy representing the inherent reliability of the model's measurement capabilities. This is particularly beneficial for effectively expanding the linear and accurate prediction range of sucrose content in peanut kernels when only a limited number of measurable peanut kernels are available. Finally, the determination coefficient R of the model is compared through cross-validation. 2 The root mean square error (RMSEP) is used to optimize the data model; the final value is the coefficient of determination R. 2＝ 0.9054, Root Mean Square Error (RMSEP) ＝ 0.6774; The final data model obtained can be packaged and applied after external verification.

[0043] Example 5

[0044] Internal validation. The chemical values ​​of sucrose content in peanut kernels from 325 standard sample sets were fitted with collected near-infrared spectral data, and a prediction model was established using the PLS method. The coefficient of determination (R²) was [not specified]. 2The value was 0.9054, and the root mean square error (RMSEP) was 0.6774. Figure 4 The model has a high coefficient of determination and a small root mean square error, indicating that the model can effectively predict the sucrose content in peanut kernels.

[0045] External validation. The near-infrared spectral predictions and chemical values ​​of sucrose content in the kernels of 20 randomly selected individuals from the RIL population of Jihua 11 × Jihua Tian 1 are shown in Table 2. The correlation scatter plot is shown below. Figure 5 .

[0046] Table 2. Validation of the near-infrared model for sucrose content in peanut kernels.

[0047]

[0048] It can be seen that its coefficient of determination (R²) 2 The value of 0.9478 indicates that the predicted sucrose content obtained by this model is relatively accurate and can be used to replace chemical determination methods for non-destructive detection of sucrose content in early generations of peanut kernels during hybridization breeding.

[0049] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0050] As can be seen from the above examples, sugar content is an important indicator affecting the edible quality and target population of peanuts, and sucrose is the main soluble sugar in mature peanuts. Establishing an efficient detection technology for sucrose content will help accelerate the improvement of high-sucrose and low-sucrose edible peanut varieties. This study utilized recombinant inbred line populations with a wide sucrose content distribution, employed near-infrared spectroscopy to collect near-infrared spectra of naturally dried peanut kernels, determined their sucrose content using high-performance liquid chromatography (HPLC), and constructed a near-infrared model of peanut kernel sucrose content using partial least squares method and its improved algorithm. The coefficient of determination (R²) of the model was [not specified in the original text]. 2 The coefficient of determination (R²) was 0.9054, and the root mean square error (RMSEP) was 0.6774. External validation of the model was performed using 20 materials. The coefficient of determination (R²) for the predicted and chemical values ​​was [not specified]. 2 The value of 0.9478 indicates that the model accurately predicts the sucrose content of peanut kernels, enabling rapid and non-destructive determination of sucrose content in individual plants during the early generations of hybridization, and effectively improving the breeding efficiency of peanut varieties with high sucrose content.

[0051] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for accurately predicting the sucrose content of peanut kernels by linearly broadening the range of a few kernels, characterized by: This method is used for early varietal trait screening in peanut breeding and / or rapid non-destructive identification of peanut kernel sweetness in agriculture / industry. It uses near-infrared spectroscopy technology, based on the sucrose content gradient data of standard samples and combined with the linear optimal algorithm of data fitting to construct a data prediction model. It further utilizes peanut generation population kernels and optimizes the data model based on observable and comparable single or multiple quantitative indicators. Under technical parameters where there are few or no peanut kernels available, it expands the accurate linear predictable range of sucrose content in peanut kernels. The data prediction model is constructed using a modified least squares method based on first-order and / or second-order differential infrared spectral data as the linear optimal algorithm for data fitting. The improved least squares method is to establish a mathematical model based on the chemical value of sucrose content and the near-infrared spectral data of peanut kernels in the standard sample set using partial least squares (PLS) based on the first-order and second-order differential infrared spectral data. The main model is constructed using the first-order differential and its derivative, and the auxiliary model is constructed using the second-order differential and its reciprocal. When fuzziness of quantitative indicators occurs under the main model during the subsequent model optimization process, the auxiliary model is referred to for auxiliary decision-making.

2. The method for accurately predicting the sucrose content of peanut kernels based on a linear expansion of the range of a few kernels according to claim 1, characterized in that: The optimization of the data prediction model is based on two quantifiable indicators, namely the coefficient of determination. R 2 The root mean square error (RMSEP) is optimized.

3. The method for accurately predicting the sucrose content of peanut kernels based on a linear expansion of a small number of kernels, as described in claim 2, is characterized in that: The optimization of the data prediction model uses the final value of its dual indicators as the coefficient of determination. R 2= 0.9054±0.004, Root Mean Square Error (RMSEP) = 0.6774±0.004; the corresponding data prediction model accurately predicts the sucrose content of peanut kernels in a linear range of ≤12% w / w, especially 7%-12% w / w.

4. The method for accurately predicting the sucrose content of peanut kernels based on a linear expansion of a small number of kernels, as described in claim 3, is characterized in that: The minimum number of usable peanut kernels is no more than 18-20.

Images