CN-121978263-A - Multi-dimensional fingerprint construction method and system for mulberry organic acid component

CN121978263ACN 121978263 ACN121978263 ACN 121978263ACN-121978263-A

Abstract

The invention belongs to the technical field of chromatographic data processing, and relates to a method and a system for constructing multidimensional fingerprints of mulberry organic acid components, comprising the following steps of preparing a mulberry sample solution containing target active ingredients by utilizing a solvent extraction technology; the method comprises the steps of establishing a standard organic acid spectrum characteristic template library, generating an associated wavelength weight coefficient table, generating a full-spectrum monitoring data stream with a trigger mark, dividing and extracting a characteristic area three-dimensional signal matrix and a non-characteristic area one-dimensional signal vector, retrieving and calling a dynamic specific weight coefficient value in the wavelength weight coefficient table based on a predicted chemical identity attribute, and generating a mulberry organic acid multidimensional fingerprint. The invention solves the problems that the existing traditional Chinese medicine fingerprint construction technology lacks the capability of dynamic integration and specific weighting of multidimensional spectrum information, and is difficult to comprehensively characterize the characteristics of complex active ingredients such as mulberry organic acid and the like.

Inventors

SHI MENG
LIU HAIYUE
WU JINNI
Shang ruibo
ZHU PAN

Assignees

陕西君碧莎制药有限公司

Dates

Publication Date: 20260505
Application Date: 20260408

Claims (10)

1. The method for constructing the multidimensional fingerprint of the mulberry organic acid component is characterized by comprising the following steps of: S1, preparing a mulberry test sample solution containing target active ingredients by using a solvent extraction technology; s2, obtaining a series of mulberry typical organic acid standard substances, detecting absorption spectrum characteristic parameters of the extracted substances through scanning, establishing a standard organic acid spectrum characteristic template library and generating a related wavelength weight coefficient table; S3, inputting the mulberry sample solution into a chromatographic flow path to execute separation operation, generating a trigger mark according to a comparison result of the light absorption intensity of the reference wavelength and a preset occurrence threshold, and restricting the multi-channel detection hardware acquisition state according to the preset occurrence threshold to generate a full-spectrum monitoring data stream with the trigger mark; S4, analyzing the full-spectrum monitoring data stream with the trigger mark, and dividing and extracting a characteristic area three-dimensional signal matrix and a non-characteristic area one-dimensional signal vector; S5, performing wavelength photoelectric information matching ratio matching optimization operation on each acquisition time section in the three-dimensional signal matrix of the characteristic region to obtain a predicted chemical identity attribute, and searching and calling out a dynamic specific weight coefficient value in a wavelength weight coefficient table based on the predicted chemical identity attribute; And S6, performing forward dimension reduction and weighting calculation on the three-dimensional signal matrix of the characteristic region by using the dynamic specific weight coefficient value to obtain a comprehensive response value sequence of the characteristic region, and performing fusion and splicing processing on the comprehensive response value sequence of the characteristic region and the one-dimensional signal vector of the non-characteristic region to generate the mulberry organic acid multidimensional fingerprint.
2. The method for constructing a multi-dimensional fingerprint of organic acid component of mulberry according to claim 1, wherein the method for preparing a sample solution of mulberry containing the target active ingredient by using a solvent extraction technique comprises the steps of: pulverizing Mori fructus pure product to a preset mesh number, adding extraction solvent, and performing ultrasonic extraction to obtain liquid phase extraction mixture; And (3) performing centrifugal impurity removal and microporous filter membrane filtration treatment on the obtained liquid phase extraction mixture to generate a clear mulberry sample solution.
3. The method for constructing the multidimensional fingerprint of the mulberry organic acid component according to claim 1, wherein the method for constructing the standard organic acid spectrum characteristic template library comprises the following steps: Injecting typical organic acid standard substances of mulberries into a chromatographic system with multi-channel detection capability, and obtaining standard full-wavelength spectrum data of each typical organic acid standard substance of mulberries; And (3) combing characteristic absorption band information in the standard full-wavelength spectrum data, extracting a spectrum profile one-dimensional luminosity numerical vector formed after vector normalization processing to generate a standard identification sequence, and constructing a standard organic acid spectrum characteristic template library by using the standard identification sequence.
4. The method for constructing a multidimensional fingerprint of mulberry organic acid composition according to claim 1, wherein the step of generating the associated wavelength weight coefficient table comprises the steps of: And organizing a representative contribution rate parameter set covering all characteristic wavelengths into a numerical value vector, carrying out normalization calculation on the representative contribution rate parameter by taking the sum of absorbance of the selected characteristic wavelength sequence as a denominator, and associating the normalized numerical value vector with the corresponding mulberry typical organic acid standard substance identity to generate a wavelength weight coefficient table.
5. The method for constructing the multidimensional fingerprint of the mulberry organic acid component according to claim 1, wherein the method for generating the full-spectrum monitoring data stream with the trigger mark comprises the following steps: Driving an infusion pump of a chromatographic system to transmit mulberry sample solution to perform chromatographic column elution separation, continuously recording the light absorption intensity of elution effluent under a preset reference wavelength parameter, and generating a reference wavelength chromatographic signal; tracking and calculating the current reading value of the reference wavelength chromatographic signal in real time, and comparing the current reading value with a preset occurrence threshold value to execute difference comparison; When the comparison judges that the current reading value is larger than a preset occurrence threshold, a channel triggering instruction is generated, the multichannel detection hardware is controlled to start a high-speed polling synchronous acquisition task for a group of characteristic wavelength sequences, and instantaneous multi-wavelength spectrum fragments with aggregation peak intensities in a retention time window are recorded; And when the comparison judges that the current reading value is not more than the preset occurrence threshold value, controlling the multichannel detection hardware to maintain basic monitoring of the reference wavelength parameter, splicing signals generated by the preamble and the instruction state record result, and merging and deriving the full-spectrum monitoring data stream with the trigger mark.
6. The method for constructing the multidimensional fingerprint of the mulberry organic acid component according to claim 1, wherein the method for dividing the three-dimensional signal matrix of the extracted characteristic region and the one-dimensional signal vector of the non-characteristic region comprises the following steps: Searching a high-intensity state time range with a trigger mark and stored with a channel trigger instruction in a full-spectrum monitoring data stream, segmenting and extracting each instantaneous multi-wavelength spectrum fragment covered by the high-intensity state time range, and reconstructing by utilizing a chromatographic retention time coordinate and a detection wavelength channel number to generate a characteristic region three-dimensional signal matrix; and screening a background baseline state time range in which a channel trigger instruction does not appear in the full-spectrum monitoring data stream with the trigger mark, extracting a cross section coordinate of the retention time and a one-to-one corresponding basic absorbance value, and constructing a non-characteristic area one-dimensional signal vector.
7. The method for constructing the multidimensional fingerprint of the mulberry organic acid component according to claim 1, wherein the dynamic specific weight coefficient value is retrieved from the wavelength weight coefficient table based on the predicted chemical identity attribute, and the method comprises the following steps: Slicing one by one along the direction of a color spectrum retention time coordinate axis to extract a detection light intensity multichannel numerical value set contained in a three-dimensional signal matrix of a characteristic region, inputting the detection light intensity multichannel numerical value set into a standard organic acid spectrum characteristic template library, and initiating similarity measurement operation with cosine; locking a target standard organic acid identifier of a first item of cosine similarity value ranking in a comparison calculation set of single slices, and taking the target standard organic acid identifier as a predicted chemical identity attribute; And searching the wavelength weight coefficient table through the chemical identity index mark according to the predicted chemical identity attribute, and extracting the corresponding dynamic specificity weight coefficient value.
8. The method for constructing the multidimensional fingerprint of the mulberry organic acid component according to claim 7, wherein the method for constructing the multidimensional fingerprint of the mulberry organic acid component is characterized in that a multi-channel numerical value set of the detected light intensity is input into a standard organic acid spectrum characteristic template library to initiate the similarity measurement operation with cosine, and comprises the following steps: slicing one by one along the direction of a color spectrum retention time coordinate axis to extract a detection light intensity multichannel numerical value set contained in the three-dimensional signal matrix of the characteristic region; preprocessing the multi-channel numerical value set of the detected light intensity to generate a query vector; And inputting the query vector into a standard organic acid spectrum characteristic template library, and calculating a cosine similarity value between the query vector and each standard identification sequence stored in the standard organic acid spectrum characteristic template library.
9. The method for constructing a multi-dimensional fingerprint of an organic acid component of mulberry according to claim 1, wherein the method for generating the multi-dimensional fingerprint of the organic acid component of mulberry comprises the steps of: transferring the read dynamic specificity weight coefficient value to apply vector cross point multiplication cooperation accumulation summation algorithm operation with each wave band independent response parameter existing in the characteristic region three-dimensional signal matrix on the same time slice section respectively, and outputting characteristic region comprehensive response value sequence; The method comprises the steps of applying a time flow synchronization algorithm tool to read the occurrence time sequence scale of absolute chromatography, embedding data point coordinates of a characteristic region comprehensive response value sequence into a non-characteristic region one-dimensional signal vector to fill an original jump gap, and completing the replacement splicing operation of a time stamp level to generate a total discontinuous sequence; And performing boundary transition alignment treatment on the overall discontinuous sequence, anchoring and leveling front and rear base points of the comprehensive response curve of the characteristic region to the base line height of the non-characteristic region, performing envelope curve re-smoothing treatment on the overall discontinuous sequence by using a spline interpolation function component to eliminate stage hard steps, and outputting the mulberry organic acid multidimensional fingerprint.
10. A multi-dimensional fingerprint construction system of mulberry organic acid components, which is applied to the multi-dimensional fingerprint construction method of the mulberry organic acid components as claimed in any one of claims 1 to 9, and is characterized by comprising the following modules: the sample solution preparation module is used for preparing a mulberry sample solution containing target active ingredients by using a solvent extraction technology; the formula dynamic analysis module acquires a series of mulberry typical organic acid standard substances, and detects absorption spectrum characteristic parameters of the extracted substances through scanning, establishes a standard organic acid spectrum characteristic template library and generates an associated wavelength weight coefficient table; The trigger type data acquisition module is used for inputting mulberry sample solution into a chromatographic flow path to execute separation operation, generating a trigger mark according to a comparison result of the light absorption intensity of a reference wavelength and a preset occurrence threshold value, and restricting a multi-channel detection hardware acquisition state according to the preset occurrence threshold value to generate a full-spectrum monitoring data stream with the trigger mark; The monitoring data flow analysis module analyzes the full-spectrum monitoring data flow with the trigger mark, and divides and extracts a three-dimensional signal matrix of the characteristic area and a one-dimensional signal vector of the non-characteristic area; The identity prediction and weight retrieval module is used for performing wavelength photoelectric information matching optimization operation on each acquisition time section in the three-dimensional signal matrix of the characteristic region to obtain a predicted chemical identity attribute, and retrieving and calling out a dynamic specific weight coefficient value in the wavelength weight coefficient table based on the predicted chemical identity attribute; and the fingerprint spectrum fusion generation module performs forward dimension reduction and weighting calculation on the three-dimensional signal matrix of the characteristic region by utilizing the dynamic specific weight coefficient value to obtain a comprehensive response value sequence of the characteristic region, and performs fusion splicing treatment on the comprehensive response value sequence of the characteristic region and the one-dimensional signal vector of the non-characteristic region to generate the mulberry organic acid multidimensional fingerprint spectrum.

Description

Multi-dimensional fingerprint construction method and system for mulberry organic acid component Technical Field The invention belongs to the technical field of chromatographic data processing, and relates to a multi-dimensional fingerprint construction method and system for mulberry organic acid components. Background The medicinal and edible traditional Chinese medicinal materials such as mulberry leaves and mulberries have complex components, the quality evaluation needs to consider the integrity and the specificity of the components, the chemical composition and the quality difference of the traditional single content measurement or the conventional detection method are difficult to comprehensively reflect, and the construction of a comprehensive fingerprint becomes a key requirement for the quality control of the traditional Chinese medicinal materials, and particularly for the active components such as mulberry organic acids, a high-efficiency fingerprint construction scheme for adapting the characteristics of the active components is needed. In order to realize comprehensive quality evaluation of traditional Chinese medicinal materials, a related fingerprint construction technical scheme has appeared in the industry, for example, the establishment of Sang Shegao performance liquid chromatography fingerprint with application number 202110651660.6 is combined with the application of multi-component content measurement in mulberry She Zhiliang evaluation, the scheme adopts specific chromatographic column and gradient elution conditions, establishes fingerprint containing 31 common peaks and identifies 6 characteristic peaks, and compensates the limitation of single component measurement in a mode of combining multi-component quantification with the fingerprint, thereby providing technical support for mulberry leaf quality evaluation. The prior art is based on single-dimension detection and fixed characteristic peak analysis of high performance liquid chromatography, can realize separation and quantification of partial components, but does not relate to integration and utilization of multidimensional spectrum information, is difficult to dynamically match chemical identities of components to be detected and carry out targeted signal weighting, and is difficult to meet quality evaluation requirements of accurate characterization of complex active components due to the fact that the components such as mulberry organic acid are similar in structure and easily overlapped in spectrum characteristics, the problems of insufficient specificity of fingerprint and limited component resolution capability are easy to occur, and deep associated information of the components cannot be mined. The traditional Chinese medicinal material fingerprint construction technology lacks dynamic integration and specific weighting capability of multidimensional spectrum information, and is difficult to comprehensively characterize complex active ingredients such as mulberry organic acid, which is a core technical problem to be solved by the invention. Disclosure of Invention In a first aspect, the invention provides a method for constructing a multidimensional fingerprint of mulberry organic acid components, which comprises the following steps: S1, preparing a mulberry test sample solution containing target active ingredients by using a solvent extraction technology; s2, obtaining a series of mulberry typical organic acid standard substances, detecting absorption spectrum characteristic parameters of the extracted substances through scanning, establishing a standard organic acid spectrum characteristic template library and generating a related wavelength weight coefficient table; S3, inputting the mulberry sample solution into a chromatographic flow path to execute separation operation, generating a trigger mark according to a comparison result of the light absorption intensity of the reference wavelength and a preset occurrence threshold, and restricting the multi-channel detection hardware acquisition state according to the preset occurrence threshold to generate a full-spectrum monitoring data stream with the trigger mark; S4, analyzing the full-spectrum monitoring data stream with the trigger mark, and dividing and extracting a characteristic area three-dimensional signal matrix and a non-characteristic area one-dimensional signal vector; S5, performing wavelength photoelectric information matching ratio matching optimization operation on each acquisition time section in the three-dimensional signal matrix of the characteristic region to obtain a predicted chemical identity attribute, and searching and calling out a dynamic specific weight coefficient value in a wavelength weight coefficient table based on the predicted chemical identity attribute; And S6, performing forward dimension reduction and weighting calculation on the three-dimensional signal matrix of the characteristic region by using the dynamic specific weight coef