CN102323285A - Method for analyzing NMR (Nuclear Magnetic Resonance) metabonomics detection data - Google Patents

Abstract

The invention provides a method for analyzing NMR (Nuclear Magnetic Resonance) metabonomics detection data and belongs to the field of biotechnology. The method mainly comprises the following processes of: step 1, obtaining a spectrogram by an NMR detection technology and processing the spectrogram; step 2, carrying out cluster analysis on a sample and drawing a sample cluster analysis column diagram; step 3, carrying out principal component analysis (PCA); step 4, carrying out partial least squares discriminant analysis (PLS-DA) and verifying a model; step 5, carrying out orthogonal partial least squares discriminant analysis (OPLS-DA); and step 6, identifying difference metabolites and carrying out correlation analysis between the metabolites. The method has the advantage of providing a novel approach for analyzing the spectrogram obtained by the NMR metabonomics detection.

Description

A kind of NMR metabolism group detects the analytical approach of data

Technical field

The invention belongs to biological technical field, relate to the data analysis aspect of the metabolism group detection of NMR (nuclear magnetic resonance), provide a kind of NMR metabolism group to detect the analytical approach of data.

Background technology

NMR (Nuclear Magnetic Resonance) is a nuclear magnetic resonance.Be the non-vanishing atomic nucleus of magnetic moment, Cai Man division, the physical process of the radio-frequency radiation of the certain frequency of resonance absorption take place in outside magnetic field effect spin energy level down.Nuclear magnetic resonance spectroscopy is a branch of spectroscopy, and its resonant frequency is in radio-frequency range, and corresponding transition is the transition of nuclear spin on nuclear Cai Man energy level.Nuclear magnetic resonance is suitable for liquid, solid.High resolution technique of today also has been used for nuclear-magnetism the research of semisolid and micro-example.Nuclear magnetic spectrogram develops into two dimension of today (2D), three-dimensional (3D) even four-dimensional (4D) spectrogram from the one dimension spectrogram (1D) in past, and they show molecular structure and intermolecular relation more clear.

Many universities, research institution and group of enterprises in the world can hear this noun of nuclear magnetic resonance, comprise the large group that we are familiar with in daily life.And it is increasingly extensive in chemical industry, oil, rubber, building materials, food, metallurgy, geology, national defence, environmental protection, weaving and other industrial sector purposes.

In China, mainly aspect fundamental research, enterprise and commercial application popularity rate are not high in its application, and main cause is that product development is not enough, use cost is higher.But use more in petrochemical complex, medical diagnostic method.The second half in 20th century; NMR technology and instrument development are very fast; From permanent magnetism to the superconduction, the magnetic field difference of the NMR spectrometer magnet from 60MHz to 800MHz improved 15 times in few per 5 years, and peculiar function promoted on organic structural analysis and medical diagnosis by NMR for this.NMR has become the analytic routines means of testing in the organic chemistry research now, and same, MRI in medical treatment (NMR imaging instrument device) also becomes the diagnostic means of some disease.

The invention provides the analytical approach that a kind of NMR metabolism group detects data.

Summary of the invention

The purpose of this invention is to provide the analytical approach that a kind of NMR metabolism group detects data.This method provides new way for the NMR metabolism group detects the data analysis that is obtained.For the characteristic of this method better is described, the present invention will explain its practical implementation process with an instantiation.This instance is to adopt rat urine, ight soil and tissue extract, carries out the NMR metabolism group and detects, and obtains original data.

Step 1, the NMR detection technique obtains spectrogram, and spectrogram is handled;

Step 2, sample cluster analysis histogram is drawn in the sample cluster analysis;

Step 3, major component (PCA) is analyzed;

Step 4, PLS, discriminatory analysis (PLS-DA) reaches verification of model;

Step 5, quadrature PLS, discriminatory analysis (OPLS-DA);

Step 6, the evaluation of difference metabolin, and correlation analysis between metabolin.

The invention has the advantages that:, a kind of new way is provided for the NMR metabolism group detects the spectrum analysis that obtains.

Description of drawings

Fig. 1.A kind of NMR metabolism group according to the invention detects the implementing procedure figure of data analysing method

Embodiment

According to the invention, a kind of NMR metabolism group detects the analytical approach of data, and the present invention adopts rat urine, ight soil and tissue extract, and carrying out the check and analysis of NMR metabolism group is instance, explains that this method gets the practical implementation step:

Step 1, the NMR detection technique obtains spectrogram, and spectrogram is handled;

(V3.0, Bruker Biospin Germany) have carried out Fourier transform to spectrogram to use TopSpin software; The phase place adjustment; Baseline correction, and calibration to wait processing, all spectrograms when carrying out Fourier transform, all to multiply by the broadening factor be that the window index function of 1Hz is with the raising signal to noise ratio (S/N ratio).

Step 2, sample cluster analysis histogram is drawn in the sample cluster analysis;

Use biostatistics software SPSS Statistics 17.0 to carry out the cluster analysis of sample, thereby the overall information of sample is carried out integral body assurance.

Step 3, major component (PCA) is analyzed;

(Umea Sweden) carries out the pattern-recognition multivariable analysis to the data after the normalization for V11.0, Umetrics AB, and the data scale conversion mode of centralization conversion (mean center scaling) is used in principal component analysis (PCA) to use SIMCA-P+ software.

Step 4, PLS, discriminatory analysis (PLS-DA) reaches verification of model;

(Umea Sweden) carries out PLS (PLS) to the data after the normalization and finds the correlationship between NMR data (X variable) and other variable (Y variable, grouping information) for V11.0, Umetrics AB to use SIMCA-P+ software.PLS-discriminatory analysis (PLS-DA) is used from the data scale conversion mode of fitting convert (unit variancescaling).PLS-DA carries out the cross validation check to the quality of model with house one method, and with R2X that obtains behind the cross validation and Q2 (the measurable degree of explainable variable of representative model and model respectively) model validity is passed judgment on.After this, through arrange experiment at random repeatedly (n=200) change putting in order of classified variable y and obtain corresponding different random Q 2 values model validity is done further check.

Step 5, quadrature PLS, discriminatory analysis (OPLS-DA).

The PLS-DA model is carried out quadrature correction process (OPLS-DA); Use SIMCA-P+ software (V11.0, Umetrics AB, Umea; Sweden) carry out quadrature PLS-discriminatory analysis (OPLS-DA), highlight the difference between the inner different groups of model substantially.OPLS-DA uses from the data scale conversion mode of fitting convert (unit variancescaling).

Step 6, the evaluation of difference metabolin, and correlation analysis between metabolin.

Through analysis,, the metabolin that statistical significance is arranged is further concluded through analyzing the corresponding related coefficient of each metabolin to OPLS-DA.In related coefficient figure, with the multiply each other conversion of recalling of laggard line data of the square root of the loading value of each variable and its standard deviation.Compare with corresponding related coefficient tables of critical values then, obtain causing the metabolin of group difference.Carry out the HPCA analysis with the various organs of control group and high dose group and the analysis integrated data of PCA of body fluid metabolic group, thereby obtain between the sample metabolin of the same race and the correlativity between the different sample metabolin.

This method gets characteristic: this method is analyzed NMR metabolism group detection data a kind of new approaches is provided.

More than be the description of this invention and non-limiting, based on other embodiment of inventive concept, all among protection scope of the present invention.

Claims (1)

1. a NMR metabolism group detects the analytical approach of data, it is characterized in that this method includes following steps:

Step 1, the NMR detection technique obtains spectrogram, and spectrogram is handled;

Step 2, sample cluster analysis histogram is drawn in the sample cluster analysis;

Step 3, major component (PCA) is analyzed;

Step 4, PLS, discriminatory analysis (PLS-DA) reaches verification of model;

Step 5, quadrature PLS, discriminatory analysis (OPLS-DA);

Step 6, the evaluation of difference metabolin, and correlation analysis between metabolin.

Images