CN109187614A - Based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method and its application - Google Patents
Based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method and its application Download PDFInfo
- Publication number
- CN109187614A CN109187614A CN201811130259.2A CN201811130259A CN109187614A CN 109187614 A CN109187614 A CN 109187614A CN 201811130259 A CN201811130259 A CN 201811130259A CN 109187614 A CN109187614 A CN 109187614A
- Authority
- CN
- China
- Prior art keywords
- data
- nuclear magnetic
- magnetic resonance
- variable
- nmr
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N24/00—Investigating or analyzing materials by the use of nuclear magnetic resonance, electron paramagnetic resonance or other spin effects
- G01N24/08—Investigating or analyzing materials by the use of nuclear magnetic resonance, electron paramagnetic resonance or other spin effects by using nuclear magnetic resonance
- G01N24/087—Structure determination of a chemical compound, e.g. of a biomolecule such as a protein
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N24/00—Investigating or analyzing materials by the use of nuclear magnetic resonance, electron paramagnetic resonance or other spin effects
- G01N24/08—Investigating or analyzing materials by the use of nuclear magnetic resonance, electron paramagnetic resonance or other spin effects by using nuclear magnetic resonance
- G01N24/088—Assessment or manipulation of a chemical or biochemical reaction, e.g. verification whether a chemical reaction occurred or whether a ligand binds to a receptor in drug screening or assessing reaction kinetics
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
- G01N30/88—Integrated analysis systems specially adapted therefor, not covered by a single one of the groups G01N30/04 - G01N30/86
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- High Energy & Nuclear Physics (AREA)
- General Physics & Mathematics (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Crystallography & Structural Chemistry (AREA)
- Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Medicinal Chemistry (AREA)
- Other Investigation Or Analysis Of Materials By Electrical Means (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method and its application, this method comprises: being utilized respectively the metabolic profile of nuclear magnetic resonance spectrometer and liquid chromatograph-mass spectrometer device acquisition biological sample, two kinds of metabolism group data of mass spectrum (MS) are obtained based on nuclear magnetic resonance (NMR) and are based on；Using the Diffreg method of inspection, the variable that there is supplement meaning to NMR data is first filtered out from MS data, the MS variables set screened is merged with NMR data collection；The variable of redundancy is deleted from new data set again, then resulting data are fused data, are convenient for subsequent modeling analysis.The model that the method for the present invention obtains has preferably explanation and estimated performance, and has very strong applicability for the data fusion of other metabolism group platforms.
Description
Technical field
The invention belongs to analysis technical field, be related to based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method and
It is applied.
Background technique
Metabolism group is one emerging " group is learned " that rear era gene occurs.It utilizes NMR spectrum and application of gas chromatorgraphy/mass
Deng high-throughput modern analytical technique, in conjunction with chemometrics method, it is qualitative fixed to carry out to the endogenous metabolites of organism
The measurement of amount ground and analysis, to disclose life entity to the response pattern of external stimulus in gene, drug and environment etc..It is organized with other and learns skill
Art such as genomics, transcription group is compared with proteomics, and metabolism group can more be truly reflected entire body a certain
Specific physiology or metabolic profile under pathological state and the vital movement having occurred and that, have become systems biology research
Another important means.
Nuclear magnetic resonance technique (NMR) and liquid chromatography mass joint technology (LC-MS) are the most frequently used in metabolism group research
Two kinds of analytical technologies.The main advantage of NMR technology is that biological sample pre-treatment step is simple, and spectrogram is reproducible and is convenient for
Quantitative analysis is metabolized the unbiased tropism of analyte detection, but sensitivity is lower.And the resolution ratio of LCMS detection and sensitivity are higher, it can be with
Detect the lower metabolin of concentration, but the reproducibility of LCMS is poor, quantitative relative difficult.Therefore, combine nuclear magnetic resonance
Can have complementary advantages with mass-spectrometric technique, obtain more fully metabolic profile information.But due to containing one in NMR and LCMS data
The information for dividing redundancy, if directly carrying out series connection modeling to the two data, these redundancies easily cause model performance decline
And it obtains and improperly analyzes result.NMR and LCMS data are how effectively merged, useful information is extracted, are to improve model performance
A critical issue.
Diffreg, which is examined, to be used to examine between two models with the presence or absence of significant difference.If the parameter of two models point
It Wei not θ(1)And θ(2), this method is for examining
The specific steps of which are as follows:
(1) data divide: it is the equal two parts of sample size by the sample random division of two models,WithWith
AndWithThe p value of the Screening and third step that are respectively used to second step calculate.
(2) process of Screening:Screening is intended to estimate a correlated variables collectionAnd it induces
Parameter Subspace outIt is assumed thatFor real correlated variables
The set of composition, then Screening process should meet two conditions: (a)Sample corresponding compared to its
Measure ninNumerical value is smaller；(b)Screening method is respectively applied toAnd pond
Change sampleObtain θ(1)、θ(2)And θ(12)EstimationWithAnd correlated variables set's
EstimationAnd
(3) p value calculates: using second part dataAnd pond sampleTo calculate p
Value.Note byWithIts log-likelihood function of each group model determined is Lind, byIts logarithm of the pond model of decision
Likelihood function is Ljoint.Consider statistic LR=2 (Lind-Ljoint), then its p can be calculated with permutation tests or its progressive distribution
Value.
Diffreg upchecks inspection: " being added after a new variables in a model, whether model occurs significantly
Variation ", to judge whether new variables has supplement meaning to existing model.In turn, if inapparent variable is considered as
Redundant variables, then Diffreg inspection can be used for deleting redundant data.Therefore, it examines, can extract by Diffreg
A kind of data have another data the information of complementarity, the redundancy in data set can also be deleted, to help
Help the model for establishing a function admirable.
Summary of the invention
It is an object of the invention to provide based on nuclear magnetic resonance and mass spectrographic metabolism in place of overcome the deficiencies in the prior art
Group learns data fusion method and its application, solves the problems in above-mentioned background technique.
The technical solution adopted by the present invention to solve the technical problems is: providing based on nuclear magnetic resonance (NMR) and mass spectrum
(MS) metabolism group data fusion method, this method efficiently extracts the supplemental information in MS to NMR data, to construct
One more good model of estimated performance.
It is of the present invention a kind of based on metabolism group data fusion method associated with NMR and MS.This method is first to sense
The experimental group and control group of interest biology carry out NMR and LC-MS respectively and detect to obtain corresponding Metabolic Fingerprinting spectrum, then pass through
Diffreg examines the additional variable group extracted in MS data relative to NMR data, and then NMR number is added in this additional variable group
According to a new growth data matrix is generated, finally this matrix is modeled.This data fusion method efficiently extracts
Complementary information in NMR and MS data set reduces interference of the redundancy to model, has great importance.
It is of the present invention a kind of based on metabolism group data fusion method, specific technical solution associated with NMR and MS
It is as follows:
(1) sample collection: collecting the laboratory sample of metabolism group research, to laboratory sample handle to obtain normal specimens A with
Disease sample B；The sample of intact animal and diseased n animal is usually collected, is properly saved in -80 DEG C of refrigerators of placement to be measured.
(2) collection and processing of data: one-dimensional nuclear magnetic resonance hydrogen spectrum is carried out to laboratory sample and liquid chromatography-mass spectrography is real
It tests, obtained spectrogram is handled to obtain every behavioral experiment sample, is often classified as the two-dimensional matrix of metabolin information, respectively
NMR data collection and MS data set；
A. nuclear magnetic resonance experiment carries out on Brooker DRX 500MHz spectrometer, and experimental temperature is set as 298K, acquisition life
The high-resolution one-dimensional nuclear magnetic resonance hydrogen of object sample is composed.
B. after sample preparation, Mass spectrometry experiments are in 1260 LC of Agilent (Agilent Technologies, Santa
Clara, CA) it carries out in AB Sciex QTrap 5500 MS (AB Sciex, Toronto, Canada) system, in positive and negative mould
Respectively detection is primary under formula, is integrated using 2.1 software of MultiQuant (AB Sciex, Toronto, Canada) to spectral peak
Processing.
C. spectrogram pre-processes: using MestReNova, v.8.1.2 software carries out Fourier to the nuclear magnetic resoance spectrum collected
The pretreatment such as variation, phase adjustment, baseline correction and calibration obtains after carrying out subsection integral and normalization to treated spectrogram
Every behavioural analysis sample, is often classified as the two-dimensional matrix of metabolin information；MzML file is first converted to the mass spectrum collected, so
File is read with R afterwards, the processing such as centralization, denoising, alignment is carried out to mass spectrum file using XCMS v3.0.2 software, is obtained every
Behavior sample is often classified as the two-dimensional matrix of metabolin information.
(3) Diffreg is selected to the p value of NMR data collection using each variable in Diffreg checking computation MS data set
Examine p value less than 0.05 or preceding 10% small variable, these variables are to have supplement to believe NMR data collection in MS data set
The variable of breath；
(4) it concentrates each variable to the p value of the additional variable collection in (3) using Diffreg checking computation NMR data, protects
Preceding 10% to 20% small variable is stayed, the NMR additional variable collection newly obtained is combined with the MS additional variable in (3), institute
The new data set obtained is nuclear magnetic resonance and mass spectrographic fused data.
Algorithm schematic diagram is as shown in Figure 1, be changed to following steps for Diffreg algorithm to examine in this scenario:
Examine screening with the variable of supplement meaning using Diffreg in the step (3), (4) method particularly includes:
1. data preparation: by the NMR data collection in (3) or the MS additional variable data set in (4) as Xlocal, will
(3) each of the NMR data collection in each of the MS data set in variable or (4) variable is as Xre, for every
One XreConstruct X=Xlocal+βXre, wherein β is XreCoefficient；Y is the label of each corresponding sample；
2. data divide: X is divided into two parts XinWith Xout, corresponding response is respectively YinWith Yout。
3. Screening: utilizing first part's data Xin, calculate separately in limitation βiModel under=0, i ∈ A with do not limit
The estimation of parameter beta in simulationWithAnd the set of correlated variablesWith
4. p value calculates: utilizing second part data Xout, calculate separately the log-likelihood function of two models
With
And calculate its likelihood ratio statistics
Equally, the p value of LR is calculated using its asymptotic distribuion or permutation tests；Wherein, the data square to be tested such as X representative
Battle array, Y represent corresponding classification, and β is to be added to the variable to be tested of matrix to be measured.
(5) redundant variables are rejected: MS additional variable group obtained above and NMR set of variables being remained, become to NMR
Each of amount group variable is a diffreg to this MS additional variable group and examines, and for example above-mentioned diffreg of specific algorithm is calculated
Method, retains in this step preceding 10% small variable in p value, and the set of variables retained is that the NMR after rejecting redundant variables becomes
Amount group.By this NMR set of variables and MS additional variable group to combination, as a sparse NMR and MS fused data matrix.
The present invention also provides application of the above-mentioned data fusion method in the building of metabolism group sparse model, make model structure
Construction method is simple, easy to promote and utilize.
In a preferred embodiment of the present invention, using of the present invention a kind of based on metabolism group associated with NMR and MS
Data fusion method, comprising the following steps: acquisition normal, disease organism biological sample to be checked carries out sample pretreatment, will be pre-
Treated, and sample carries out nuclear magnetic resonance, mass spectroscopy, obtains the original metabolism spectrum of sample to be examined；By these original metabolism spectrums into
The pretreatment of row spectrogram, obtains the two-dimensional data matrix comprising metabolin information；Obtained data matrix is used for this metabolism group
Data fusion method, available fused data matrix, modeling, which is further analyzed, to this data matrix can be used for disease
The multiple-tasks such as prediction or classification.
The technical program compared with the background art, it has the following advantages:
1. the biological sample that biological sample of the invention is normal bio and diseased organism (can be blood, urine, excrement
Deng), such biological sample is common sample survey, collects and treatment process is simple and convenient, and cost is low, is answered convenient for promoting
With.
2. by the way that the information from NMR and MS can be integrated based on metabolism group data fusion method associated with NMR and MS,
A new fused data matrix is established, facilitates and continues analysis and modeling in next step.This method is for improving pathological diagnosis and dividing
The accuracy rate of type has biggish help, and the proposition of this method is so far at home and abroad there is not yet similar report.
3. metabolism group data fusion method of the invention, it can be not only used for nuclear magnetic resonance and mass spectrum both detection skills
Art Metabolic Fingerprinting obtained composes convergence analysis, can also extend to nuclear magnetic resonance and near infrared spectrum, mass spectrum and near infrared light
The Model Fusion of other analysis platforms such as spectrum.
4. being usually used in combination nuclear magnetic resonance and when mass-spectrometric technique is modeled, due to variable is various, dimension explosion etc. because
Element will cause the difficulty in analysis；For this method in by fusion nuclear magnetic resonance information and Information in Mass Spectra, construction one is relatively sparse
Variables set, this variables set remains the information useful to model construction, while eliminating the information of redundancy, to subsequent carry out mould
Type building, the work such as prediction have very great help.
Detailed description of the invention
Fig. 1 is algorithm flow chart of the invention.
Fig. 2 is additional variable figure of the single MS variable to NMR data matrix.
Fig. 3 is that NMR and MS binds directly the shot chart for being PLSDA.
Fig. 4 is that the fusion matrix after extracting complementarity does the shot chart of PLSDA.
Fig. 5 is the CV verifying of the PLSDA of the fusion matrix after extracting complementarity.
Specific embodiment
Embodiment 1
1. research object
The present invention uses 83 serum samples acquired in state of Indiana University Medical College, respectively from 28 Colon and rectums
Cancer patient and 55 healthy volunteers, all colorectal cancer patients are to be diagnosed for the first time, are not carried out before blood sampling any
Operative treatment, chemotherapy and radiotherapy.
2. reagent and instrument and equipment
Experiment reagent includes heavy water, chloroform, methanol, sodium dihydrogen phosphate, dipotassium hydrogen phosphate, ammonium acetate, acetic acid.
Testing consumptive material includes 5mm nuclear magnetic tube, pipettor, 1000 μ L and 200 μ L pipette tips, 5mL EP pipe, marking pen, latex hand
Set and mask.
Experimental facilities includes pipettor, low-temperature and high-speed centrifuge, Brooker 500MHz nuclear magnetic resonance spectrometer, Agilent 1260
5500 MS system of LC and AB Sciex QTrap.
3. sample collection and pretreatment
Acquire the blood plasma of 28 colorectal cancer patients and 55 healthy volunteers, collected blood stand 45 minutes after with
2000 revs/min of revolving speed is centrifuged 10 minutes, it takes supernatant to be stored in -80 DEG C of refrigerators until experiment after centrifugation.
4. nuclear magnetic resonance spectroscopy
530 μ L samples to be tested are taken to be added in 5mm NMR pipe.By a 60 μ L TSP of encapsulation (3-propionic-2,2,3,
3-d4 acid sodium salt) (20.9nmol TSP is dissolved in 60 μ L D to solution2In O) capillary be placed in 5mm NMR pipe
It is interior to calibrate (δ=0.00ppm) for chemical shift.1D1H NMR experiment carries out on Bruker DRX 500MHz spectrometer.In advance
Sequence is saturated to inhibit water peak signal and acquire in conjunction with CPMG (Carr-Purcell-Meiboom-Gill) spin-echo sequence
Serum Spectrum.Test temperature is set as 25 DEG C, each spectrum adds up 128 times altogether, and spectrum width 6kHz, collected FID signal carries out Fu
Spectrogram signal-to-noise ratio is improved multiplied by an exponential weighting function before vertical leaf transformation.Utilize Bruker TopSpin software (version
3.5) phase and baseline correction are carried out to spectrogram.It according to spectrum peak position and splits point information metabolin is belonged to, and to acquisition
Metabolin spectral peak integrated manually.
5.LC/MS spectrum measurement
50 μ L serum samples and 150 μ L methanol vortex mixeds, and it is statically placed in -20 DEG C of refrigerators 20 minutes；Take out sample with
14000 revs/min of speed is centrifuged 10 minutes, takes supernatant；For remaining Protein pellets, 300 μ L first are added again
Alcohol is centrifuged 10 minutes for vortex mixed 10 minutes, 14000 revs/min, takes supernatant later；The supernatant mixing extracted in two-wheeled
After be dried；Solid after drying is redissolved in 99.8% ammonium acetate solution (5mM ammonium acetate, 40% water/60% second of 500 μ L
Nitrile)+0.2% acetic acid (5.13 μM of l-tyrosine-13C2/ 22.54 μM of Pfansteihl sodium-13C3), two of them isotope labelling
Reference substance is used for monitoring experiment system stability.LC-MS/MS is tested in 1260 LC (Agilent of Agilent
Technologies, Santa Clara, CA) and 5500 MS of AB Sciex QTrap (AB Sciex, Toronto, Canada)
It is carried out in system.Respectively detection is primary under each positive and negative mode of sample.Target mass spectrometric data is in mass spectrum multiple-reaction monitoring
It is obtained under (multiple reaction monitoring, MRM) mode.Detect 59 pairs and 99 respectively under positive and negative mode
A parent-daughter ion is to (in total 158 pairs).Finally, using 2.1 software of MultiQuant (AB Sciex, Toronto,
Canada Integral Processing) is carried out to spectral peak.
It is established 6. supplemental information is extracted with fused data matrix
Data matrix importing Rstudio software is analyzed, after all data carry out centralization processing, it will be handled
Nuclear magnetic data matrix afterwards be labeled as NMR, mass spectrometric data matrix be labeled as MS, first to each of MS variable respectively with
NMR does Diffreg inspection after combining, and before p value 10% small variable is extracted, as MS additional variable collection；Then right
Each of NMR variable does Diffreg inspection with MS additional variable collection set respectively, same to extract preceding 10% small variable,
As NMR additional variable collection, as soon as in conjunction with the two additional variable collection as a new data matrix, this new data matrix
It is by sparse NMR and MS fused data matrix.
7. method validation and comparison
It is better than to verify the performance of modeling of the extracted data fusion matrix of this method in step 6 and binds directly NMR
With the performance of modeling of the data matrix of MS, we to the data matrix in step 6 and can bind directly the data of NMR and MS
Matrix does PLSDA analysis, and PLSDA algorithm can call directly the algorithm in SIMCA software.After obtaining two PLSDA models,
Shot chart can be drawn to compare: shot chart is as shown in Figure 4.PLSDA analysis (Fig. 3) comparison is done with NMR and MS is directly combined,
It can be seen that the contribution rate of the first two principal component is improved close to 20%, and it can see this Fusion Model and passed through CV verifying
(Fig. 5).For the ease of comparing, we carry out PLSDA modeling to NMR data, MS data and its tandem data respectively, carry out pair
Than as a result such as table one.In Table 1 it can be seen that, when two kinds of data are directly used in conjunction, the PMSE (predict of PLSDA model
Mean-square error predicts mean square error) not only compared to not declined with MS data individually, risen instead, but
It is to be extracted MS modeling and forecasting mean square error is smaller, and performance is better to combining after the complementarity variable of NMR.
The prediction result of the PLSDA model of 1. different data of table compares
7. conclusion
It is of the present invention a kind of based on metabolism group data fusion associated with NMR and MS it can be seen from verifying above
Fusion matrix constructed by method can be improved the accuracy and predictability of subsequent modeling analysis, compared to directly using nuclear-magnetism it is total
Vibration spectrum or mass spectrum modeling, new method shows certain superiority, and efficiently utilizes mass spectrometric data to nuclear magnetic resonance
Supplemental information in data.
Embodiment 2
Using a kind of metabolism group sparse model construction method based on multi-platform data fusion of the present invention, including
Following steps:
(1) biological sample to be checked is acquired according to above-mentioned steps 3 to be pre-processed, reach nuclear magnetic resonance experiment and Mass spectrometry experiments
It is required that；
(2) pretreated biological sample to be checked is subjected to nuclear magnetic resonance spectroscopy according to above-mentioned steps 4, obtains life to be checked
These original metabolic profile maps are carried out spectrogram pretreatment, obtained comprising metabolin by the original metabolic profile map of object sample
The nuclear magnetic resonance two-dimensional data matrix of information；
(3) pretreated biological sample to be checked is subjected to mass spectroscopy according to above-mentioned steps 5, obtains biological sample to be checked
Mass spectrum；These raw mass spectrums are subjected to spectrogram pretreatment, obtain the mass spectrum two-dimensional data matrix comprising metabolin information；
(4) two data matrixes are imported into based on metabolism group data fusion associated with NMR and MS according to above-mentioned steps 6
Method generates fused data matrix.
The above is only the preferred embodiment of the present invention, the range implemented of the present invention that therefore, it cannot be limited according to, i.e., according to
Equivalent changes and modifications made by the invention patent range and description, should still be within the scope of the present invention.
Claims (6)
1. being based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method, which comprises the steps of:
(1) sample collection: the laboratory sample of metabolism group research is collected, laboratory sample is handled to obtain normal specimens A and disease
Sample B；
(2) collection and processing of data: carrying out one-dimensional nuclear magnetic resonance hydrogen spectrum to laboratory sample and liquid chromatography-mass spectrography tested, will
Obtained spectrogram is handled the two-dimensional matrix for obtaining every behavioral experiment sample, being often classified as metabolin information, respectively NMR data
Collection and MS data set；
(3) Diffreg is selected to examine p the p value of NMR data collection using each variable in Diffreg checking computation MS data set
It is worth less than 0.05 or preceding 10% small variable, these variables are to have supplemental information to NMR data collection in MS data set
Variable；
(4) concentrate each variable to the p value of the additional variable collection in (3) using Diffreg checking computation NMR data, before reservation
10% to 20% small variable combines the NMR additional variable collection newly obtained with the MS additional variable in (3), resulting
New data set is nuclear magnetic resonance and mass spectrographic fused data.
2. according to claim 1 be based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method, it is characterised in that:
The experiment of one-dimensional nuclear magnetic resonance hydrogen spectrum is completed on Brooker DRX 500MHz spectrometer in the step (2), and experimental temperature is set as
298K acquires the one-dimensional nuclear magnetic resonance hydrogen spectrum of laboratory sample；Liquid chromatography-mass spectrography is tested in 1260 LC AB of Agilent
It is completed in 5500 MS system of Sciex QTrap, is respectively detected under each positive negative mode of laboratory sample and once obtain its metabolic information.
3. according to claim 1 be based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method, it is characterised in that:
The spectral data of the spectrum of one-dimensional nuclear magnetic resonance hydrogen described in the step (2), is calibrated, phase school using MestReNova software
Just, baseline correction, spectral peak alignment, delete remaining solvent peak and baseline point, then carry out subsection integral and normalized, obtain
NMR data collection；The spectral data of liquid chromatography-mass spectrography integrates spectral peak using MultiQuant, then uses QC sample
Data are normalized, MS data set is obtained.
4. according to claim 1 be based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method, it is characterised in that:
Examine screening with the variable of supplement meaning using Diffreg in the step (3), (4) method particularly includes:
1. data preparation: by the NMR data collection in (3) or the MS additional variable data set in (4) as Xlocal, will be in (3)
Each of MS data set variable or (4) in each of NMR data collection variable as Xre, for each
XreConstruct X=Xlocal+βXre, wherein β is XreCoefficient；Y is the label of each corresponding sample；
2. data divide: X is divided into two parts XinWith Xout, corresponding response is respectively YinWith Yout；
3. Screening: utilizing first part's data Xin, calculate separately in limitation βiModel under=0, i ∈ A and mould is not limited
The estimation of parameter beta in typeWithAnd the set of correlated variablesWith
4. p value calculates: utilizing second part data Xout, calculate separately the log-likelihood function of two models
With
And calculate its likelihood ratio statistics
Equally, the p value of LR is calculated using its asymptotic distribuion or permutation tests；Wherein, the data matrix to be tested such as X representative, Y generation
The corresponding classification of table, β are to be added to the variable to be tested of matrix to be measured.
5. according to claim 4 be based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method, it is characterised in that:
The redundant variables for deleting that growth data is concentrated are examined in the step (4) using Diffreg method particularly includes: retain
Preceding 10% small variable in p value in Diffreg algorithm.
6. such as application of any one of the Claims 1 to 5 metabolism group data fusion method in the building of metabolism group sparse model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811130259.2A CN109187614B (en) | 2018-09-27 | 2018-09-27 | Metabonomics data fusion method based on nuclear magnetic resonance and mass spectrum and application thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811130259.2A CN109187614B (en) | 2018-09-27 | 2018-09-27 | Metabonomics data fusion method based on nuclear magnetic resonance and mass spectrum and application thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109187614A true CN109187614A (en) | 2019-01-11 |
| CN109187614B CN109187614B (en) | 2020-03-06 |
Family
ID=64907443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811130259.2A Expired - Fee Related CN109187614B (en) | 2018-09-27 | 2018-09-27 | Metabonomics data fusion method based on nuclear magnetic resonance and mass spectrum and application thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109187614B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109632860A (en) * | 2019-01-15 | 2019-04-16 | 中国科学院昆明植物研究所 | A kind of method of monomeric compound structure in parsing mixture |
| CN109917061A (en) * | 2019-03-11 | 2019-06-21 | 嘉兴迈维代谢生物科技有限公司 | A kind of method of quick point in wide target metabolism group |
| CN111370067A (en) * | 2020-02-28 | 2020-07-03 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | LC/GC-MS-oriented metabonomics data quality control method and system |
| CN111474202A (en) * | 2020-04-09 | 2020-07-31 | 集美大学 | Internal standard liquid for analyzing metabolites in biological body fluid and application thereof |
| CN113049664A (en) * | 2021-03-15 | 2021-06-29 | 东华理工大学 | Path analysis modeling method based on mass spectrum metabonomics |
| CN114527154A (en) * | 2022-02-09 | 2022-05-24 | 国仪量子(合肥)技术有限公司 | Method for metabolomics sample component detection |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002086501A2 (en) * | 2001-04-23 | 2002-10-31 | Metabometrix Limited | Methods for analysis of spectral data and their applications: osteoporosis |
| WO2003017177A2 (en) * | 2001-08-13 | 2003-02-27 | Beyong Genomics, Inc. | Method and system for profiling biological systems |
| WO2004038602A1 (en) * | 2002-10-24 | 2004-05-06 | Warner-Lambert Company, Llc | Integrated spectral data processing, data mining, and modeling system for use in diverse screening and biomarker discovery applications |
| CN102100706A (en) * | 2009-12-18 | 2011-06-22 | 中国科学院大连化学物理研究所 | Method for evaluating quality of Chinese patent medicament by using metabonomics |
| CN103604875A (en) * | 2013-06-08 | 2014-02-26 | 江苏警官学院 | Method for measuring serum metabolism markers in methylamphetamine abusers |
| CN104713970A (en) * | 2015-04-01 | 2015-06-17 | 山东省肿瘤医院 | Method for constructing serum metabonomics analysis model |
| CN104713971A (en) * | 2015-04-01 | 2015-06-17 | 山东省肿瘤医院 | Method for analyzing serum metabolomics on basis of LC-MS (liquid chromatogram-mass spectrograph) serum metabolomics technology |
| CN107133448A (en) * | 2017-04-10 | 2017-09-05 | 温州医科大学 | A kind of metabolism group data fusion optimized treatment method |
-
2018
- 2018-09-27 CN CN201811130259.2A patent/CN109187614B/en not_active Expired - Fee Related
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002086501A2 (en) * | 2001-04-23 | 2002-10-31 | Metabometrix Limited | Methods for analysis of spectral data and their applications: osteoporosis |
| WO2003017177A2 (en) * | 2001-08-13 | 2003-02-27 | Beyong Genomics, Inc. | Method and system for profiling biological systems |
| WO2004038602A1 (en) * | 2002-10-24 | 2004-05-06 | Warner-Lambert Company, Llc | Integrated spectral data processing, data mining, and modeling system for use in diverse screening and biomarker discovery applications |
| CN102100706A (en) * | 2009-12-18 | 2011-06-22 | 中国科学院大连化学物理研究所 | Method for evaluating quality of Chinese patent medicament by using metabonomics |
| CN103604875A (en) * | 2013-06-08 | 2014-02-26 | 江苏警官学院 | Method for measuring serum metabolism markers in methylamphetamine abusers |
| CN104713970A (en) * | 2015-04-01 | 2015-06-17 | 山东省肿瘤医院 | Method for constructing serum metabonomics analysis model |
| CN104713971A (en) * | 2015-04-01 | 2015-06-17 | 山东省肿瘤医院 | Method for analyzing serum metabolomics on basis of LC-MS (liquid chromatogram-mass spectrograph) serum metabolomics technology |
| CN107133448A (en) * | 2017-04-10 | 2017-09-05 | 温州医科大学 | A kind of metabolism group data fusion optimized treatment method |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109632860A (en) * | 2019-01-15 | 2019-04-16 | 中国科学院昆明植物研究所 | A kind of method of monomeric compound structure in parsing mixture |
| CN109917061A (en) * | 2019-03-11 | 2019-06-21 | 嘉兴迈维代谢生物科技有限公司 | A kind of method of quick point in wide target metabolism group |
| CN111370067A (en) * | 2020-02-28 | 2020-07-03 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | LC/GC-MS-oriented metabonomics data quality control method and system |
| CN111370067B (en) * | 2020-02-28 | 2023-05-23 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | LC/GC-MS-oriented metabonomics data quality control method and system |
| CN111474202A (en) * | 2020-04-09 | 2020-07-31 | 集美大学 | Internal standard liquid for analyzing metabolites in biological body fluid and application thereof |
| CN113049664A (en) * | 2021-03-15 | 2021-06-29 | 东华理工大学 | Path analysis modeling method based on mass spectrum metabonomics |
| CN114527154A (en) * | 2022-02-09 | 2022-05-24 | 国仪量子(合肥)技术有限公司 | Method for metabolomics sample component detection |
| WO2023151577A1 (en) * | 2022-02-09 | 2023-08-17 | 国仪量子(合肥)技术有限公司 | Metabonomics sample composition detection method |
| CN114527154B (en) * | 2022-02-09 | 2024-02-02 | 国仪量子技术(合肥)股份有限公司 | Method for detecting components of metabonomic sample |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109187614B (en) | 2020-03-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109187614A (en) | Based on nuclear magnetic resonance and mass spectrographic metabolism group data fusion method and its application | |
| Wishart | Quantitative metabolomics using NMR | |
| Barding et al. | Quantitative NMR for bioanalysis and metabolomics | |
| Dunn et al. | Measuring the metabolome: current analytical technologies | |
| Wu et al. | Determination of total concentration of chemically labeled metabolites as a means of metabolome sample normalization and sample loading optimization in mass spectrometry-based metabolomics | |
| Patti et al. | Metabolomics: the apogee of the omics trilogy | |
| Mercier et al. | Towards automatic metabolomic profiling of high-resolution one-dimensional proton NMR spectra | |
| Schripsema | Application of NMR in plant metabolomics: techniques, problems and prospects | |
| Zhang et al. | Modern analytical techniques in metabolomics analysis | |
| CN104777242B (en) | Associating mark, test kit and system for diagnosis of polycystic ovary syndrome | |
| US20060200316A1 (en) | Data correction, normalization and validation for quantitative high-throughput metabolomic profiling | |
| CN109564207A (en) | For detecting and the mass spectrometry method of Quantitative metabolite object | |
| CN106018600B (en) | A kind of metabolism group method for distinguishing false positive mass spectrum peak-to-peak signal and quantitative correction mass spectrum peak area | |
| CN113777209B (en) | Synchronous detection and application of exposure and effect markers of volatile pollutants in urine | |
| Beasley‐Green et al. | A proteomics performance standard to support measurement quality in proteomics | |
| CN111579665A (en) | UPLC/HRMS-based metabonomics relative quantitative analysis method | |
| Djukovic et al. | Mass spectrometry and NMR spectroscopy based quantitative metabolomics | |
| CN109791158A (en) | The method that more attributes for complex sample monitor | |
| Banoei et al. | Metabolomics in critical care medicine: a new approach to biomarker discovery | |
| Bowling et al. | Analyzing the metabolome | |
| Jones et al. | An introduction to metabolomics and its potential application in veterinary science | |
| Bjerrum | Metabonomics: analytical techniques and associated chemometrics at a glance | |
| Liu et al. | GC/TOFMS analysis of endogenous metabolites in mouse fibroblast cells and its application in TiO 2 nanoparticle-induced cytotoxicity study | |
| Hoyt et al. | Perspective: A potential role for NUS in metabolite‐based in vitro diagnostics. | |
| Kotłowska | Application of steroid hormone metabolomics in search of biomarkers in clinical research |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200306 Termination date: 20200927 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |