CN108398491A - A kind of quality determining method of Dendrobium loddigesii - Google Patents

Abstract

The invention discloses a quality detection method of Dendrobium meihua, including: 1) using ITS-26SE and ITS-17SE as primers to perform sequencing to identify the species of the Dendrobium medicinal material to be tested; 2) performing chromatographic detection on samples with a sample capacity of n , to obtain the detection data of chemical small molecule components schaftoside and/or naringenin as a reference component; 3) carry out fingerprint spectrum detection on samples respectively, and obtain the peak area value of the fingerprint spectrum of the full chemical composition of Dendrobium candidum; 4 ) Use the component content value of Schaftoside and/or naringenin in the chromatographic data as the response variable, and use the peak area value of other components in the fingerprint as the independent variable to establish an analysis model and establish a chemical small molecule component by screening variables with the Lasso method The relevant feature fingerprint model. Through generation sequencing and characteristic fingerprints, the quality of Dendrobium candidum can be accurately identified and controlled.

Description

A kind of quality detection method of Dendrobium candidum

technical field

The invention belongs to the field of detection and analysis of components of traditional Chinese medicines, and in particular relates to a quality detection method of Dendrobium candidum.

Background technique

Dendrobium loddigesii Rolfe.: also known as Dendrobium loddigesii Rolfe. Stems are weak, often drooping, thin cylindrical, 10cm-30cm long, 0.2cm-0.4cm in diameter, sometimes branched, with many nodes, internodes 1.5cm-2cm long, golden yellow after drying; leaf bases with sheaths, leaf sheaths Quality, sheath mouth often open after drying. Raceme laterally borne on the upper part of the old stem with leaves, with 1-2 flowers; inflorescence peduncle 0.2cm-0.3cm long, with 1-2 short, cup-shaped membranous sheaths at the base; flowers white or purple Red; bracts are membranous, ovate, 0.2cm long, with blunt apex; pedicel and ovary are light green, 2cm-3cm long; middle sepals are ovate-oblong, 1.7cm-2cm long, 0.6cm-0.7cm wide, apex obtuse Acute, with 5 veins; lateral sepals lanceolate, 1.7cm-2cm long, 0.6cm-0.7cm wide, apex acute, with 5 veins, base oblique; calyx capsule nearly spherical, about 0.5cm long; petals Elliptic, as long as middle sepal, 0.8cm-0.9cm wide, apex slightly obtuse, entire, with 3-5 veins; lip nearly round, 1.7cm-2cm in diameter, upper center golden yellow, periphery pale Purple red, slightly concave, with short fringes on the edge, densely pubescent on both sides; stamen white, with red stripes on both sides of the front, about 0.4cm long; medicine cap white, nearly conical, densely covered with fine papillary hairs, front edge With irregular teeth. The stem of Dendrobium candidum can be used as medicine, sweet and slightly cold. Tonify stomach and promote body fluid, nourish yin and clear away heat. It is used for yin deficiency and body fluid deficiency, dry mouth, polydipsia, retching due to lack of food, asthenia and heat after illness, and unknown vision.

Dendrobium has always been regarded as a precious Chinese herbal medicine, which has very important nourishing effects. Dendrobium is clinically used to treat a variety of diseases, and has pharmacological effects such as enhancing immunity, anti-oxidation, lowering blood sugar and inhibiting cancer. Due to the long-term unrestrained excavation and irrational use of Dendrobium, its wild resources are decreasing day by day, and some phenomena of false ones and inferior ones have appeared in the market. In addition, due to the large number of Dendrobium species, the hybridization between the species leads to the crossover phenomenon of the characters of its close relatives, and it is difficult to classify and distinguish. Therefore, it is necessary to establish the characteristic fingerprint of Dendrobium to evaluate the medicinal quality of Dendrobium.

Chromatographic fingerprinting is a comprehensive and quantifiable means of identification. As a panoramic mode of global analysis, it reflects the overall situation of the sample. However, in the process of chromatographic fingerprint analysis, a lot of data are high-dimensional, that is, the data contains many attributes or characteristics, such as the chromatographic fingerprint of Dendrobium candidum, which can better describe Dendrobium beautiful, but in practical applications Directly operating high-dimensional data in the medium will face the problem of "curse of dimensionality", which will lead to an exponential increase in the number of samples required for the modeling process as the dimensionality increases. In the face of high-dimensional data, the conventional least squares method is no longer applicable. In order to improve the interpretability of the model and the accuracy of prediction, variable selection becomes very important. How to efficiently screen out several variables that play an important role in the dependent variable from numerous variables is an urgent problem to be solved when analyzing fingerprints.

At present, the National Pharmacopoeia uses the quantitative determination of the content of a certain active ingredient or active ingredient, that is, small molecule ingredients in Chinese medicinal materials, to evaluate its quality. However, studies have proved that the curative effect of traditional Chinese medicine comes from the synergistic effect between its various "active ingredients", and even the generally recognized effective "active ingredient" and "inactive ingredient". "In order to achieve the curative effect of traditional Chinese medicine, rather than the result of a single active ingredient. In traditional Chinese medicine under the guidance of the theory of traditional Chinese medicine, any active ingredient cannot fully reflect the overall curative effect of traditional Chinese medicine.

Contents of the invention

The invention provides a quality detection method of Dendrobium candidum, which uses first-generation sequencing to identify the species of Dendrobium medicinal materials, and uses the Lasso method to screen variables to establish a fingerprint model of the relevant characteristics of chemical small molecule components in Dendrobium candidum. The characteristic fingerprint model accurately evaluates the quality of medicinal materials of Dendrobium meihua.

The purpose of the present invention is achieved through the following technical solutions:

A quality detection method of Dendrobium meihua, comprising:

1) Take ITS-26SE: 5'GAATTCCCCGGTTCGCTCGCCGTTAC 3';

ITS-17SE: 5'ACGAATTCATGGTCCGGTGAAGTGTTCG 3' is used as a primer, and PCR amplification and sequencing are performed to identify the species of Dendrobium to be tested as a sample of Dendrobium meihua;

2) Carry out chromatographic detection on the Dendrobium officinale sample with a sample size of n, and obtain the detection data with chemical small molecule components schaftoside and/or naringenin as reference components;

3) carry out fingerprint spectrum detection to sample respectively, obtain the fingerprint spectrum peak area value of the full chemical composition of Dendrobium candidum;

4) With the component content value of Schaftoside and/or naringenin in the chromatographic data as the response variable, the peak area value of other components in the fingerprint spectrum is used as the independent variable to establish the analysis model, through Lasso (The Least AbsoluteShrinkage and Selection Operator) method to screen variables to establish the relevant feature fingerprint model of chemical small molecule components, the basic model of which is:

y=X ^T β+ε

Among them, y is the response variable, y=(y ₁ ,y ₂ ,...,y _n ) ^T ; X is the matrix, X=(x ₁ ,x ₂ ,...,x _n ) ^T ; E(ε )=0; Var(ε)=σ ² I _n ; ε is the random error item of the model; σ is the standard deviation of the random error item; n is the sample size; I _n is an n×n unit matrix.

Assume that the random terms obey the classical assumptions, namely:

(1) The random item has zero mean, E(ε _i | _xi )=0;

(2) Random items have the same variance, Var(ε _i | _xi )=σ ² ;

(3) Random items have no serial correlation, Cov(ε _i ,ε _j )＝0, i≠j;

(4) ε follows normal distribution, ε _i ~N(0,σ ² ).

The variance matrix of the random entries is a square matrix with ^σ2 on the diagonal and zeros elsewhere, as follows:

Among them, I _n is an n×n unit matrix, n is the sample size of the data,

Further, the Lasso method is realized by formula I calculation:

In Formula I, n is the sample size; p ^* is the number of variables; p is the dimension of the sample; y=(y ₁ ,y ₂ ,...,y _n ) ^T ∈ R ⁿ is the response variable; x=( x ₁ ,x ₂ ,...,x _n ) ^T is an n×p design matrix, including all candidate independent variables that affect the response variable; λ is an adjustment parameter; is the penalty function; the meaning of β ₀ is the intercept term of the formula, that is, the value of the response variable y when all independent variables x are 0; the meaning of β _j is the coefficient of the independent variable x _j , that is, the effect of the independent variable x _j on the response The degree of influence of the variable y.

Further, the selection method of λ is the K-fold cross-validation method:

K-fold CVs:

Wherein, K is 5 or 10.

Further, the analysis model is a sub-model with the smallest CV value.

Further, the selection of λ follows the GCV criterion, and the GCV criterion is defined as:

Among them, SSE _k is the residual sum of squares of the CV submodel containing k variables, df=trace{P(λ)}; trace represents the trace of the matrix. In linear algebra, the sum of the elements on the main diagonal (diagonal from upper left to lower right) of an n×n matrix A is called the trace (or trace number) of matrix A, generally denoted as tr(A). That is, df is equal to the sum of all elements on the main diagonal in the matrix P(λ).

Further, the analysis model is a sub-model with the smallest GCV value.

Further, when the chromatographic data presents an ultra-high-dimensional situation, the following SIS (Sure Independence Screening) method is first used to screen variables, and then the Lasso method is used for processing;

SIS: M _γ ＝{1≤i≤p:|ω _i | is the first |γn| relatively large}

Among them, M _* ={1≤i≤p:β _i ≠0} indicates the subscript set of non-zero coefficients in the true model; s=|M _* | indicates the number of non-zero coefficients; ω=(ω ₁ ,ω ₂ ,...,ω _p ) ^T ＝X ^T y; for any given γ∈(0,1), the p elements of ω are arranged and defined in descending order of absolute value; at this time |γn|< n, select the independent variable corresponding to the subscript in M _γ , which is to reduce the ultra-high dimension to d (d≤n) dimension; where, d=n or d=[n/log n].

The present invention also provides the application of the above method in the quality control of Dendrobium meihua.

Compared with the prior art, the present invention has at least the following advantages:

(a) In the present invention, first use ITS-26SE and ITS-17SE as primers to measure the characteristic sequence of the Dendrobium medicinal material, so as to determine that the Dendrobium medicinal material is the species of Dendrobium meihua; The data is used as an independent variable, and the chromatographic data of other components in the above-mentioned detection data is used as a dependent variable to establish a linear regression model of small molecule components and fingerprints, so that the quality evaluation of Dendrobium medicinal materials is more accurate;

(b) The present invention adopts the Lasso method to carry out variable selection on the fingerprint of Dendrobium meihua, which effectively solves the problem of "curse of dimensionality";

(c) The present invention reduces the dimensionality of the original fingerprints, and establishes the relevant characteristic fingerprints of the chemical small molecule components of Dendrobium candidum, which is more pertinent and applicable to the interpretation of the content of a single chemical small molecule component;

(d) The present invention realizes the correlation analysis of the content of the chemical small molecule components through the relevant characteristic fingerprints of the chemical small molecule components of Dendrobium meihua, and can effectively identify and control the quality of the medicinal materials of Dendrobium meihua;

(e) When the chromatographic data presents an ultra-high-dimensional situation, the SIS method is used to reduce the dimension first, and then the Lasso method is used for processing.

Description of drawings

Fig. 1 is the fingerprint chromatogram of the whole composition of Dendrobium meihua;

Fig. 2 is the chromatogram of Schaftoside chemical small molecular substance standard sample;

Fig. 3 is the chromatogram of naringenin chemical small molecular substance standard sample;

Fig. 4 is the chromatogram of Schaftoside chemical small molecules in Dendrobium candidum;

Fig. 5 is a chromatogram of small chemical molecules of naringenin in Dendrobium candidum.

Note: No. 1 peak Schaftoside; No. 2 peak naringenin.

Detailed ways

The present invention will be described in further detail below in conjunction with the accompanying drawings and examples. The following examples are only descriptive, not restrictive, and cannot limit the protection scope of the present invention.

The generation sequencing of embodiment 1 American flower dendrobium

First-generation sequencing primer sequences:

ITS-26SE: 5'GAATTCCCCGGTTCGCTCGCCGTTAC 3';

ITS-17SE: 5' ACGAATTCATGGTCCGGTGAAGTGTTCG 3'.

The amplification and sequencing parameters are: denaturation at 98°C for 2 minutes and then PCR cycle, the PCR cycle parameters are 98°C for 20s; 52°C for 30s; 68°C for 1min, 38 cycles, 68°C for 7min, set 4°C after the amplification, and perform Generation Molecular Sequencing.

Through the first-generation sequencing, the species of Dendrobium to be tested was identified as Dendrobium meihua.

The extraction method of embodiment 2 American flower dendrobium

Take the dried sample of Dendrobium candidum, pulverize it with a pulverizer, pass through a pharmacopoeia sieve (aperture 0.335mm), accurately weigh 1.000g of Dendrobium powder (the weighing error cannot exceed 0.2%), place it in a 100ml conical flask, and add 50mL of 75 % methanol (V water: V methanol = 25:75), take it out after ultrasonication for 30 minutes at room temperature, filter, and concentrate the filtrate to dryness by rotary evaporation, dissolve it with 75% methanol solvent (V water: V methanol = 25:75), and finally transfer Dilute to a 10ml volumetric flask, shake well, and filter with a 0.45 μm microporous membrane to obtain a sample solution of Dendrobium candidum.

The chromatographic detection method of the Dendrobium officinale extract of embodiment 3

①Preparation of reference solution

Accurately weigh 4.10 mg of Schaftoside and 4.08 mg of naringenin, place them in 10 ml volumetric flasks, add 75% (V/V) methanol to dissolve and dilute, shake well, and use them as stock solutions. Refrigerate at 4°C for later use.

Then accurately draw a certain amount of reference substance stock solution and dilute with 75% methanol to accurately prepare a mixed reference substance solution of schaffertoside and naringenin. Through different dilution ratios, 7 concentration points of the prepared components were diluted. into a high performance liquid chromatograph.

②Extraction and processing method of samples for determination of small molecule components in Dendrobium meihua:

Take 1.00g of the powder of this product (passed through a No. 3 sieve), accurately weigh it, put it in a 100ml volumetric flask, add 50ml of methanol-water (75:25) accurately, ultrasonicate (power 250W, frequency 40kHz) for 30 minutes, let cool, After filtration, the filtrate was concentrated to dryness by rotary evaporation, dissolved in 5ml of methanol-water (75:25), the supernatant was passed through a 0.45μm microporous membrane, and the subsequent filtrate was obtained.

③ Chromatographic conditions for determination of small molecule components in Dendrobium meihua:

Chromatographic conditions:

Chromatographic conditions for content determination: Grace Allitima C18 chromatographic column (250mm*4.6mm, 5μm); mobile phase adopts binary gradient elution system, phase A: 0.2% acetic acid-water, phase B: acetonitrile; gradient elution program is shown in Table 1 ; Determination of naringenin with a wavelength of 290nm, and determination of schaffertoside with a wavelength of 334nm; the reference wavelength is 500nm, the column temperature is 30°C; the flow rate is 1.0mL/min, and the injection volume is 20μL.

Fingerprint chromatography conditions: Grace Allitima C18 chromatographic column, preferably 250mm×4.6mm, 5μm chromatographic column; mobile phase: A phase: 0.4% acetic acid + 20mmol/L ammonium acetate aqueous solution, B phase: acetonitrile; gradient elution: 0～12min: 2%～15% Phase B, 12～35min: 15%～24% Phase B, 35～45min: 24%～36% Phase B, 45～60min: 36%～75% Phase B, 60～ 80min: 75%-95% phase B; flow rate 1.0mL/min; column temperature 30°C; injection volume 20μL; detection wavelength 280nm.

Table 1 The elution gradient for the determination of small molecule components in Dendrobium meihua

Fig. 1 is the fingerprint chromatogram of the whole composition of Dendrobium candidum, detection wavelength 280nm;

Fig. 2 is the detection spectrum figure of Schaftoside (No. 1 peak) standard sample;

Fig. 3 is the detection spectrum figure of naringenin (No. 2 peak) standard sample;

Fig. 4 is the detection spectrum of schaftoside (No. 1 peak) in Dendrobium candidum;

Figure 5 is the detection spectrum of naringenin (Peak No. 2) in Dendrobium candidum.

Embodiment 4 Establishment of the correlation characteristic fingerprint spectrum with schaffertoside in Dendrobium candidum

1. Dendrobium sample solution preparation

Take the dry sample of Dendrobium, pulverize it with a pulverizer, pass through a pharmacopoeia sieve (aperture 0.335mm), accurately weigh 1.000g of Dendrobium powder (the weighing error cannot exceed 0.2%), place it in a 100ml Erlenmeyer flask, add 50mL of 75% methanol ( V water: V methanol = 25:75), take it out after ultrasonication for 30 minutes at room temperature, filter, and the filtrate is concentrated to dryness by rotary evaporation, dissolved in 75% methanol solvent (V water: V methanol = 25:75), and finally transferred to a capacity of 10ml Make up the volume in the bottle, shake well, and filter with 0.45μm microporous membrane to obtain the product. Table 2 is the obtained linear relationship of the reference substance schaftoside.

Table 2 The linear relationship table of the reference substance Schaftoside

2. Establishment method of the relevant characteristic fingerprint of Dendrobium meihua schaftoside

The first step: calculate the correlation coefficient of all covariates x and y;

Step 2: Arrange the absolute values of the correlation coefficients from large to small, and select the first 2√n covariates, denoted as x_1, x_2,...,x_p;

The third step: perform linear regression on y and x_1, x_2,..., x_p, and use Lasso method to select variables.

The first part of the Lasso (Least Absolute Shrinkage and Selection Operator) function indicates the goodness of the model fit, and the second part can be regarded as a penalty. This method compresses small coefficients toward 0. Once a coefficient is compressed to 0, the corresponding variable is deleted. It's like filtering with a "sieve", sifting out variables that have a small impact once. The smaller the λ, the more variables in the model. The larger the λ, the greater the shrinkage, and the fewer variables are selected. The Lasso method is a continuous and orderly process with small variance. When the adjustment parameter is large enough, the penalty term has the effect of forcing the estimated value of some of the coefficients to be set to 0, so the Lasso method can perform variable selection and obtain a sparse model.

When the independent variable is p, the sample size is n, and when p>>n, the SIS method is used to reduce the dimension first, and then the Lasso method is used to screen variables.

SIS M _γ ＝{1≤i≤p:|ω _i |is the first |γn| relatively large ones}

Among them, M _* ={1≤i≤p:β _i ≠0} indicates the subscript set of non-zero coefficients in the true model; s=|M _* | indicates the number of non-zero coefficients; ω=(ω ₁ ,ω ₂ ,...,ω _p ) ^T ＝X ^T y; for any given γ∈(0,1), the p elements of ω are arranged and defined in descending order of absolute value; at this time |γn|< n, select the independent variable corresponding to the subscript in M _γ to reduce the superhigh dimension to d(d≤n) dimension; where, d=n or d=[n/log n].

The linear model screened by the Lasso method is:

Among them, y _i is the i-th response variable, y=(y ₁ ,y ₂ ,...,y _n )'; _Xi is a covariate of order P _n ×1, X=(x ₁ ,x ₂ , ..., x _n )'; ε _i is the random error item of iid with mean value 0 and variance σ ² , E(ε)=0, Var(ε)=σ ² I _n .

Assume that the random terms obey the classical assumptions, namely:

(1) The random item has zero mean, E(ε _i | _xi )=0;

(2) Random items have the same variance, Var(ε _i | _xi )=σ ² ;

(3) Random items have no serial correlation, Cov(ε _i ,ε _j )＝0, i≠j;

(4) ε follows normal distribution, ε _i ~N(0,σ ² ).

The variance matrix of the random entries is a square matrix with ^σ2 on the diagonal and zeros elsewhere, as follows:

Among them, I _n is an n×n unit matrix, n is the sample size of the data,

In order to perform variable selection and parameter estimation at the same time, the Lasso method is realized by the minimization of the penalty least squares objective function (I).

Wherein, y=(y ₁ ,y ₂ ,...,y _n ) ^T ∈ R ⁿ is the response variable vector. Taking the dendrobium data as an example, each species of Dendrobium has two response variable sequences (schaffertoside (μg/g), naringenin (μg/g)). The response variable is affected by the independent variable, and generally y is a continuous variable.

x=(x ₁ ,x ₂ ,...,x _n ) ^T is an n×p design matrix, including all candidate independent variables that affect the response variable.

p is the dimension of the sample, and n is the sample size. In the Dendrobium data, the dimension p is much larger than the sample size n, so the least squares estimation is no longer applicable, and the method of variable selection is needed to estimate the model.

β=(β ₁ ,β ₂ ,...,β _p ) ^T is a p-dimensional parameter.

is a penalty function, and λ is an adjustment parameter. In the variable selection method, the balance between the excellent degree of model fitting and the punishment for the number of selected variables is reflected by different criteria, and here it is realized by directly selecting the adjustment parameters. Different λ values correspond to different the intensity of punishment. The larger the λ, the stronger the degree of compression, and the fewer non-zero parameters are finally estimated. The most common method for selecting λ is the K-fold cross-validation method:

K-fold CVs:

Generally, K may be 5 or 10.

The GCV criterion is an approximate situation when K in the CV criterion is taken as n, which is defined as:

Among them, SSE _k is the residual sum of squares of the CV submodel with k variables, df=trace{P(λ)}. For the selection of the final optimal model, the sub-model with the smallest CV value or GCV value can be selected.

Using the linear model, since the initial independent variable p=434, the sample size n=13, p>>n, so variable screening is a relatively important task. After dimensionality reduction by SIS, variables are screened by Lasso method.

There are 4 independent variables screened by Lasso method, and the R-square is 0.8725.

The independent variables and their corresponding coefficients screened by the Lasso method are shown in Table 3. The first column is the number of the selected independent variables, the second column is the corresponding coefficient, the third column is the variance of the coefficient, and the fourth column is the test P value.

Table 3 Included independent variables and their corresponding coefficients obtained by Lasso method

The results in Table 3 show that the screening results with Schaftoside as the response variable: the first column is the variables selected by the Lasso method, that is, the variables 241, 148, 183, and 383 are selected, and the corresponding p values ( Column 4) are all less than the significance level of 0.05, which means there is a significant difference. The meaning of the above independent variable: the peak area value of the fingerprints of the n batches of samples (n not less than 10) of Dendrobium meihua that were aligned according to the retention time.

Column 2 is the β parameter value corresponding to each variable. A positive β value indicates that the variable has a positive impact on the small molecule of Dendrobium schiaftoside; a negative β value indicates that the variable has a negative impact on the small molecule of Dendrobium schiaftoside. The absolute value of the β value shows the degree of influence of this variable on the small molecule of Dendrobium schiaftoside. Specifically, in Table 3, the impact of variable 148 on the small molecule of Dendrobium Schafertoside is positive; the impact of variables 241, 183, and 383 on the small molecule of Dendrobium Schafertoside is negative, and the negative impact of variable 183 is larger.

Example 5 Establishment of the relevant characteristic fingerprint of Dendrobium meihua naringenin

1. Preparation of reference solution

Accurately weigh 4.08mg of naringenin, put it in a 10ml volumetric flask, add 75% methanol to dissolve and dilute, shake well, and use it as a stock solution. Refrigerate at 4°C for later use. Then accurately draw a certain amount of reference substance stock solution, add 75% methanol to dilute, and accurately prepare naringenin reference substance solution. Through different dilution ratios, the prepared components were diluted to 7 concentration points and injected into a high performance liquid chromatograph. Table 4 is the linear relationship obtained for the reference substance naringenin.

Table 4 Reference substance naringenin linear relationship table

2. Establishment of naringenin-related feature fingerprints

Using the linear model, since the initial independent variable p=434, the sample size n=17, p>>n, so variable screening is a relatively important task. The present invention adopts Lasso method to screen variables.

There are 4 independent variables screened by Lasso method, and the R-square is 0.7399.

The independent variables and their corresponding coefficients screened by the Lasso method are shown in Table 5. The first column is the number of the selected independent variables, the second column is the corresponding coefficient, the third column is the coefficient variance, and the fourth column is the test P value.

Table 5 The selected independent variables and their corresponding coefficients obtained by the Lasso method

The results in Table 5 show that with naringenin as the independent variable result: the variable selected by the Lasso method is shown in column 1, that is, variables 262, 324, 421, and 392 have been screened out, and the p-value of the variable (column 4 ) are less than the significance level of 0.05, which means there is a significant difference. The meaning of the above independent variable: the peak area value of the fingerprints of n batches of samples (n not less than 10) of Dendrobium candidum after being aligned according to the retention time.

Column 2 gives the β parameter values corresponding to each variable. A positive β value indicates that the variable has a positive impact on the small molecule of Dendrobium naringenin; a negative β value indicates that the variable has a negative impact on the small molecule of Dendrobium naringenin. The absolute value of the β value shows the degree of influence of this variable on the Dendrobium naringenin small molecule. Specifically, in Table 5, variables 324, 421, and 392 have positive effects on small molecules of Dendrobium naringenin, among which variable 421 has a greater impact on small molecules of Dendrobium naringenin; variable 262 has a small effect on Dendrobium naringenin. The effect of the numerator is negative.

According to the results of the first-generation molecular data, this technology can clearly distinguish Dendrobium japonica from other Dendrobium species, and has the function of identifying species. The molecular sequence of the first-generation sequencing can be used as an identification index for this species. At the same time, the content of two chemical small molecules in the characteristic fingerprint of Dendrobium meihua is stable in each sample of the species, and can be clearly distinguished from other Dendrobium, which can also be used as one of the identification indicators of the species. Therefore, both the first-generation data and the fingerprints can be used as indicators for identifying the species, and both have the following correlations: (1) when the first-generation data identifies the sample as Dendrobium americanum, the fingerprint of the sample has specific characteristics, that is, by The species can be identified by one generation of data, and the fingerprint characteristics of this species can be known; (2) When the fingerprint identifies the sample as Dendrobium candidum, its first generation data can also be inferred. Therefore, the two are sufficient and necessary conditions for each other, and since the fingerprint characteristics of this species are stable in each representative sample, the content of the medicinal ingredients of this species can be determined. Therefore, the first-generation sequencing data and fingerprints can be used as identification indicators for this species, and can also be used as evaluation indicators for medicinal materials. If the first-generation sequencing and fingerprints are used in the variety identification and quality evaluation, the identification results of the two methods can be combined to make the identification results of Dendrobium medicinal materials more accurate and reliable.

Example 7 Application of first-generation sequencing of Dendrobium meihua and related feature fingerprints of small molecule components in quality evaluation of Dendrobium medicinal materials

1. Use the next generation sequencing primer sequence and amplification sequencing parameters to determine the sequence of the Dendrobium medicinal material to be tested, and identify it as the Dendrobium medicinal material.

First-generation sequencing primer sequences:

ITS-26SE: 5'GAATTCCCCGGTTCGCTCGCCGTTAC 3';

ITS-17SE: 5'ACGAATTCATGGTCCGGTGAAGTGTTCG 3';

The amplification and sequencing parameters are: denaturation at 98°C for 2 minutes and PCR cycle, the PCR cycle parameters are 98°C for 20s; 52°C for 30s; 68°C for 1min, 38 cycles, 68°C for 7min, and set 4°C to keep warm after amplification.

Through the above sequencing, it was determined that the Dendrobium medicinal material to be tested was the species of Dendrobium meihua.

2. Preparation of samples for testing

Accurately weigh the Dendrobium meihua powder, put it in a 100ml volumetric flask, add 50ml of methanol-water with a volume ratio of 75:25 for each 1g sample, and ultrasonically treat it with 250W power and 40kHz frequency for 30 minutes, filter after cooling, and spin-evaporate the filtrate Concentrate to dryness, dissolve every 1g of Dendrobium meihua powder with 5ml of methanol-water with a volume ratio of 75:25, pass the supernatant through a 0.45μm microporous membrane, and take the subsequent filtrate to obtain the small molecular components of Dendrobium meihua Assay samples.

3. Chromatographic detection

Chromatographic conditions:

Chromatographic column: Grace Allitima C18 chromatographic column (250mm×4.6mm, 5μm); mobile phase: A phase: 0.4% acetic acid + 20mmol/L ammonium acetate aqueous solution, B phase: acetonitrile; Gradient elution: 0～12min, 2%～ 15%B, 12～35min, 15%～24%B, 35～45min, 24%～36%B; 45～60min, 36%～75%B; 60～80min, 75%～95%B, flow rate 1.0 mL/min; column temperature 30°C; injection volume 20μL; detection wavelength 280nm.

Sample preparation method:

Take 1.00g of the powder of this product (passed through a No. 3 sieve), accurately weigh it, put it in a 100ml volumetric flask, add 50ml of methanol-water (75:25) accurately, ultrasonicate (power 250W, frequency 40kHz) for 30 minutes, let cool, After filtration, the filtrate was concentrated to dryness by rotary evaporation, dissolved in 5ml of methanol-water (75:25), the supernatant was passed through a 0.45μm microporous membrane, and the subsequent filtrate was obtained.

When detecting, measure the full-component fingerprint chromatogram with a wavelength of 280nm, and compare the similarity between the obtained full-component fingerprint and the fingerprint shown in Figure 1; the similarity greater than 0.85 is qualified.

Chinese medicine fingerprint is a variety of quantitative features that reflect the distribution of complex chemical substances contained in Chinese medicinal materials, semi-finished products and Chinese patent medicines (or herbal medicines) detected by analytical instruments. The type, quantity and content characteristics of chemical components contained in semi-finished products and Chinese patent medicines (or botanical medicines), and quantify the map that reveals potentially complex biological activity information characteristics.

The present invention first determines the species of Dendrobium through first-generation sequencing, and then conducts correlation research by measuring the content data of the small molecule components of Dendrobium and the entire peak area of Dendrobium fingerprints, and finds out the internal correlation of Dendrobium through correlation data modeling, so that comprehensive Evaluate the quality of Dendrobium, and can effectively and accurately identify and control the quality of the medicinal materials of Dendrobium meihua, making the analysis results more reliable and avoiding the interference of other types of Dendrobium medicinal materials.

The above is only a preferred embodiment of the present invention, but the scope of protection of the invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention are all Should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (7)

1. a kind of quality determining method of Dendrobium loddigesii, which is characterized in that including：

1) with ITS-26SE：5 ' GAATTCCCCGGTTCGCTCGCCGTTAC 3 ' and

ITS-17SE：5 ' ACGAATTCATGGTCCGGTGAAGTGTTCG 3 ' are primer, carry out PCR amplification sequencing, are waited for identification The kind for surveying stem of noble dendrobium medicinal material is Dendrobium loddigesii；

2) chromatography detection is carried out to the sample that sample size is n, obtained with chemical small molecule ingredient schaftoside and/or naringenin As the detection data with reference to ingredient；

3) finger-print detection is carried out respectively to sample, obtains the finger-print peak area value of the full chemistry ingredient of Dendrobium loddigesii；

It 4), will be in finger-print with the component content value of schaftoside in chromatographic data and/or naringenin variable in response The peak area value of other ingredients establishes analysis model as independent variable, passes through Lasso (The Least Absolute Shrinkage and Selection Operator) method selection variables establish the correlated characteristic fingerprint of chemical small molecule ingredient Spectrum model, basic model are：

Y=X^Tβ+ε

Wherein, y is response variable, y=(y₁,y₂,...,y_n)^T；X is matrix, X=(x₁,x₂,...,x_n)^T；E (ε)=0；Var (ε)=σ²I_n；ε is the stochastic error of model；σ is the standard deviation of stochastic error；N is sample size；I_nIt is the list of a n × n Position battle array.

2. according to the method described in claim 1, it is characterized in that, the Lasso methods are to calculate to realize by formula I：

In formula I, n is sample size；p^*For variable number；P is the dimension of sample；Y=(y₁,y₂,...,y_n)^T∈RⁿBecome for response Amount；X=(x₁,x₂,...,x_n)^TFor the design matrix of n × p, including all candidate independents variable influential on response variable；λ is Adjusting parameter；For penalty；β₀Meaning be formula intercept item, that is, when all independent variable x be 0 when respond The value of variable y；β_jIt is meant that independent variable x_jCoefficient, i.e. independent variable x_jTo the influence degree of response variable y.

3. according to the method described in claim 2, it is characterized in that, the selection method of the λ, which is K, rolls over cross-validation method：

K-fold CV：

Wherein, K is 5 or 10.

4. according to the method described in claim 3, it is characterized in that, the analysis model is to take the submodel of CV values minimum.

5. according to the method described in claim 2, it is characterized in that, the selection of the λ follows GCV criterion, the GCV criterion are fixed Justice is：

Wherein, SSE_kIt is the residual sum of squares (RSS) of the CV submodels containing k variable, df=trace { P (λ) }.

6. according to the method described in claim 5, it is characterized in that, the analysis model is to take the submodel of GCV values minimum.

7. according to the method described in claim 2, when chromatographic data is presented superelevation and ties up situation, following SIS (Sure are used first Independence Screening) method selection variables, recycle the processing of Lasso methods；

SIS：M_γ={ 1≤i≤p:|ω_i| before being | γ n | a bigger

Wherein, M_*={ 1≤i≤p:β_i≠ 0 } the subscript collection of nonzero coefficient in true mode is indicated；S=| M_*| indicate nonzero coefficient Number；ω=(ω₁,ω₂,...,ω_p)^T=X^Ty；For any given γ ∈ (0,1), the p element of ω by absolute value from It to minispread and defines greatly；At this time | γ n | ＜ n choose M_γThe corresponding independent variable of middle subscript makes superelevation dimension drop to d (d≤n) Dimension；Wherein, d=n or d=[n/log n].