CN104267108A - Chinese herbal medicinal ingredient analysis method based on combination of MREM (multi-reconstruction entropy minimization) stoichiometry and chromatography - Google Patents

Abstract

The invention relates to a Chinese herbal medicinal ingredient analysis method based on the combination of MREM (multi-reconstruction entropy minimization) stoichiometry and chromatography. The method comprises the following steps: (1) performing singular value decomposition on a matrix to obtain a VI matrix, and finding out a target peak or a target interval; (3) randomly generating a unit vector T and carrying out VT calculation to obtain a reconstruction spectrum, and unitizing the reconstruction spectrum; (4) establishing a minimum target function and a corresponding penalty function by utilizing Shannon entropy on the basis of the step 3; (5) optimizing T by utilizing an optimization method, repeating the steps 3 to 5 until the optimization is completed; and (6) determining the spectrogram (step 3) corresponding to the optimal T to be a final reconstruction spectrogram, comparing the final reconstruction spectrogram with the existing database to obtain the structure information of a substance represented by the reconstruction spectrum. Compared with the prior art, not only can the requirement on the spectrum separation be lowered, and the analysis speed be higher, but also the method is a universal Chinese herbal medicinal ingredient analysis method and applicable to a majority of traditional Chinese medicinal varieties.

Description

A kind of based on MREM stechiometry and chromatograph joint used traditional Chinese medicine ingredients analytical approach

Technical field

The present invention relates to traditional Chinese medicine ingredients method for separating and analyzing, specifically refer to that a kind of pure color based on MREM stechiometry and each composition of chromatograph joint used analysis gas phase Chinese medicine is composed and mass spectrum, belong to Pharmaceutical Analysis field.

Background technology

Chinese medicine is the precious legacy of China's medicine and pharmacology, for a long time to the procreation of the Chinese nation with powerfully made indelible contribution.Traditional Chinese medicine research for a long time ignores the further investigation to material foundation of tcm, particularly to the basic research of main pharmacodynamics material in traditional Chinese medicine complex system, lack new thinking and effective method, substantially cannot illustrate the mechanism of action of Chinese medicine from molecular level, quantitatively cannot hold the leading role of effective substance in Chinese medicine.Form so historied reason of situation, but main cause is that traditional Chinese medicine ingredients analysis difficulty is large.

Analysis of Chinese Traditional Medicine is difficult, and main cause is that the composition of Chinese medicine is very complicated, and major part all belongs to unknown component.Mass spectrometry is very capable, but is but helpless for potpourri.Chromatographic process is strong separation means, and chromatogram and mass spectrometer coupling are become one of conventional method analyzing traditional Chinese medicine ingredients.But due to traditional Chinese medicine ingredients very complicated, be difficult to make all the components all reach the effect of baseline separation under certain chromatographic condition, if directly analyzed the spectrum peak of overlap, very large error will be produced to the result of qualitative, quantitative, even can obtain the result of mistake.

Stechiometry is an emerging subject in mid-term 1970's, the instrument of this subject applied mathematics, statistics and computer science and means design or select optimum chemical method for measurement, and by resolving chemical metric data to obtain chemistry and relevant information to greatest extent, be the powerful of resolving complicated unknown system.

MREM is the abbreviation of multi-reconstruction entropy minimization, and the method is the mathematical method based on principle of least entropy producing.Entropy is a value weighing confusion, and famous principle of entropy increase is: entropy always trends towards increasing, and namely trends towards confusion.The method utilizes Mathematics Optimization Method exactly, finds out the minimum entropy of system, namely the most orderly in system state, so just pure spectrum can be separated from mixed spectrum.

Summary of the invention

The object of the present invention is to provide a kind of method solving overlap peak problem in traditional Chinese medicine sample gaschromatographic mass spectrometry figure, application MREM stechiometry, pure color spectrum and the mass spectrum of each component is told from the spectrum peak of overlap, thus the requirement reduced chromatographic resolution, improve the accuracy of Chinese medicine qualitative and quantitative analysis.

The object of the present invention is achieved like this: based on MREM stechiometry and chromatograph joint used traditional Chinese medicine ingredients analytical approach, it is characterized in that, adopts gas chromatograph-mass spectrometer, computing machine and MREM stechiometry to carry out; Concrete steps comprise:

1) Traditional Chinese drug mixture spectrogram is exported to a two-dimentional matrix;

2) then svd (SVD) is done to this matrix, obtain a V ^tmatrix;

3) by observing V ^tmatrix, finds interested peak or interval, and this step is called at " lock peak ";

4) stochastic generation vector of unit length T, then calculates with VT, obtains one and rebuilds pure spectrum (being also a vector), and utilizes the peak or interval that pin, and it is unitization to rebuild pure spectrum;

5) based on the pure spectrum of reconstruction after unitization, build one utilize Shannon entropy minimize objective function and corresponding penalty function;

6) then use an optimization method, constantly optimize T, and repeat step 4-6 until optimized;

7) last, the spectrogram (step 4) corresponding with this optimal T is exactly the final pure spectrum of reconstruction.

8) pure spectrum and database comparison (if there is) will be rebuild, obtain the structural information etc. of material representated by the pure spectrum of this reconstruction.

Compared to existing technology, the present invention has following beneficial effect:

1, the requirement to chromatographic resolution is reduced.Because the method can resolve overlap peak, each component of sample does not need to be separated completely, only needs the variable concentrations gradient obtaining different component by chromatogram, reduces the requirement to chromatographic resolution.

2, analysis speed is fast, cost-saving.Because this technology does not need the research carrying out chromatogram analysis method, the requirement to instrument and equipment in modern Chinese herbal medicine analysis and research can be saved in a large number, and the process of condition optimizing can be reduced, shorten analysis time, reduce analysis cost and the analysis time of Chinese medicine.

3, the present invention is a kind of general Chinese medicine isolation technics.Modern Methods, for a set of analytical approach of a certain Chinese medicine, often can not be used for the analysis of another kind of Chinese medicine, and in the present invention, the requirement of chromatogram only be needed to obtain concentration gradient, actual separation is not needed to open, so can be suitable for most of herbal species.

Embodiment

Below in conjunction with the drawings and specific embodiments, the invention will be further described.

Based on MREM stechiometry and a chromatograph joint used traditional Chinese medicine ingredients analytical approach, gas chromatograph-mass spectrometer, computing machine and MREM stechiometry is adopted to carry out.Fig. 1 is analysis process schematic diagram.

1 data acquisition.

Gas chromatograph-mass spectrometer is used to analyze traditional Chinese medicine sample.Instrument uses the prompt 7890A gas chromatography/5975C GC-MS of peace.Sample is Rosa Damascana, is injected directly into GC and tests, and the raw data of instrument is exported as net CDF form.Fig. 2 a is its total ion current figure.

2. rebuild pure spectrum

2.1 data conversions

Fig. 2 b is the TIC spectrogram of sample retention time in 1687-1696 second.Due to complicated component, a lot of component cannot be separated completely.Retention time two-dimentional mass spectrometric data of potpourri spectrogram between 1687-1696 second derived, the mass-to-charge ratio of each mass spectrometric data is from 57-183m/z, and sampling interval is 1m/z.Peak containing an overlap peak bunch A, via expert system process, extracts, obtains hybrid matrix A by data _{k × v}.Wherein k is computer sampling number of times, such as, between appearance time 28.17-28.27 minute, every sampling in 0.1 second, samples 90 times altogether.υ is data channel number, and being such as sampled mass spectrographic mass-charge ratio (m/z) is 57-183, and interval is 1 m/z, totally 126 data channel.

2.2 svd

By data set A _{k × u}svd (SVD) is carried out, then by right singular matrix V according to equation (1) ^tv _{× v}the part of physical significance is not had to remove, by diagonal matrix S _{k × v}in full null part remove, obtain V ^t _{k × v}and S _{k × k}, matrix U does not use.

$A_{k\×\mathrm{\υ}}=U_{k\×k}\×S_{k\×\mathrm{\υ}}\×V_{\mathrm{\υ}\×\mathrm{\υ}}^T---\left(1\right)$

To matrix V ^t _{k × v}in each row vector observe, the row vector of ground unrest removed, such as, after jth row vector, seem to be all noise, we just remove (j+1)-k row vector, obtain V ^t _{j × v.}

2.3 lock peaks

From the V that those remain ^t _{j × v}in, by inspecting row vector, user may to some or several peak, or is interested in a certain interval.Such as peak m/z=91, or m/z=90-100 is interested, just by these peaks or interval locking (Target), and can start computing.

2.4MREM

Find out a pure spectrum with optimization point-score, and calculate corresponding pseudo-concentration.Concrete steps are as follows:

First, a random vector T _{1 ' j}by Practical computer teaching, this T vector upgrades by used optimized algorithm, automatically until find optimal value.By T, a vectorial a relevant with pure spectrum ^estobtained by equation (2).A ^estchange according to T is changed, terminates equation until optimize; The a finally obtained ^estit is exactly pure spectrum.

${a^{\mathrm{est}}}_{1\×\mathrm{\υ}}=T_{1\×j}\×(S_{j\×j}\×V_{j\×\mathrm{\υ}}^T)---\left(2\right)$

By a ^estwith the top in the peak locked or interval, unitization according to equation (3).The process of the vector after unitization is exactly " lock peak " process.Wherein, a ' and a " be the peak and interval that lock.

${\hat{a}}_{1\×\mathrm{\υ}}=\frac{{a_{1\×\mathrm{\υ}}}^{\mathrm{est}}}{\max \left(a^{\′}\right)+\max \left(a^{\′\′}\right)}---\left(3\right)$

The a obtained utilized _hatbuild objective function and the penalty function of optimization method, equation (4) is total target and method function, and wherein P function is penalty function.By equation (5), calculate pseudo-concentration information.

$\min \left(G\right)=\underset{\mathrm{\υ}}{\mathrm{\Σ}}{\hat{a}}_{1\×\mathrm{\υ}}+P({\hat{a}}_{1\×\mathrm{\υ}},{\hat{c}}_{k\×1})---\left(4\right)$

${\hat{c}}_{k\×1}=A_{k\×\mathrm{\υ}}\×{\hat{a}}_{\mathrm{\υ}\×1}^T\×({\hat{a}}_{\mathrm{\υ}\×1}\×{\hat{a}}_{\mathrm{\υ}\×1}^T)---\left(5\right)$

Repeat above-mentioned steps, optimize T, with expert system comparison, remove the pure spectrum that overlaps, obtain a pure spectrum collection.

Utilize BREM method, by pinning peak m/z=57 and 183, the pure spectrum that reconstruction obtains is as Fig. 3.

3-is with database comparison.

By the NIST comparing carried with Agilent GC-MS instrument, obtain chemical formula and the structural formula of each pure spectrum.To lock the pure spectrum that peak 57 and 183 obtains, what confirm the pure spectrum representative of this reconstruction is 6,10-Dodecadien-1-ol, 3,7,11-trimethyl-(3,7,11-trimethyl 6,10-12 carbon diene-1-alcohol) and α-Bisabolol (bisabolol), comparison the results are shown in Figure 4.

4. rebuild the concentration of pure spectrum

Existing method is generally utilize TIC area, is undertaken quantitatively by internal standard method or external standard method.As can be seen from Figure 3, total TIC concentration (blue dotted line) that rebuilds shows good consistance with experiment TIC concentration (black line).(total rebuild TIC concentration be the TIC concentration of rebuilding pure spectrum simply adding with.) total peak area of total peak area and original collection of illustrative plates of rebuilding pure spectrum reach 98.4% consistance, and peak type is also corresponding.The pure spectrum of reconstruction due to us can compare with experiment TIC area, so also can utilize existing GC-MS quantivative approach, carries out quantitative test to the pure spectrum of various reconstruction.

Accompanying drawing illustrates:

Fig. 1 is based on MREM stechiometry and chromatograph joint used traditional Chinese medicine ingredients analytical approach process flow diagram

The total ion current figure of Fig. 2 a sample

Fig. 2 b sample retention time is at the total ion current figure of 1687-1696 second

Fig. 3 locks peak m/z=57 and 183, the pure spectrum of the reconstruction calculated with EM

Fig. 4 locks the matching result of pure spectrum that peak m/z=57 and 183 obtains and NIST database.

Claims (1)

1. a traditional Chinese medicine ingredients analytical approach, is characterized in that, adopts gas chromatograph-mass spectrometer, computing machine and MREM stechiometry to carry out; The method comprises the following steps:

1) use gas chromatograph-mass spectrometer to analyze traditional Chinese medicine sample, Traditional Chinese drug mixture spectrogram is exported to a two-dimentional matrix A _{k × v}, wherein k is computer sampling number of times, and υ is data channel number;

2) by matrix A _{k × v}svd (SVD) is carried out, then by right singular matrix V according to equation (1) ^t _{v × v}the part of physical significance is not had to remove, by diagonal matrix S _{k × v}in full null part remove, obtain V ^t _{k × v}and S ^{k × k}, matrix U does not use;

To matrix V ^t _{k × v}in each row vector observe, the row vector of ground unrest removed, such as, after jth row vector, seem to be all noise, we just remove (j+1)-k row vector, obtain V ^t _{j × v}.

3) V is found ^tinterested peak or interval in matrix;

4) by Practical computer teaching random vector T _{1 ' j}, this T vector upgrades by used optimized algorithm, automatically until find optimal value.By T, a vectorial a relevant with pure spectrum ^estobtained by equation (2).A ^estchange according to T is changed, terminates equation until optimize; The a finally obtained ^estit is exactly pure spectrum;

5) by a ^estwith the top in the peak locked or interval, unitization according to equation (3).The process of the vector after unitization is exactly " lock peak " process.Wherein, a ' and a " be the peak and interval that lock;

6) a obtained utilized _hatbuild objective function and the penalty function of optimization method, equation (4) is total target and method function, and wherein P function is penalty function.By equation (5), calculate pseudo-concentration information;

7) repeat above-mentioned steps, optimize T, with expert system comparison, remove the pure spectrum that overlaps, obtain a pure spectrum collection;

8) by and the NIST comparing that carries of gas chromatograph-mass spectrometer, obtain chemical formula and the structural formula of each pure spectrum;

9) utilize TIC area, carry out quantitative test by internal standard method or external standard method.