CN105938521A - Ankylosing spondylitis forewarning model establishing method and device - Google Patents
Ankylosing spondylitis forewarning model establishing method and device Download PDFInfo
- Publication number
- CN105938521A CN105938521A CN201610516917.6A CN201610516917A CN105938521A CN 105938521 A CN105938521 A CN 105938521A CN 201610516917 A CN201610516917 A CN 201610516917A CN 105938521 A CN105938521 A CN 105938521A
- Authority
- CN
- China
- Prior art keywords
- ankylosing spondylitis
- early
- data
- warning
- warning model
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B25/00—ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- Databases & Information Systems (AREA)
- Biomedical Technology (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Genetics & Genomics (AREA)
- Biophysics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biotechnology (AREA)
- Evolutionary Biology (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Theoretical Computer Science (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
The embodiment of the invention discloses an ankylosing spondylitis forewarning model establishing method and device. The method comprises the steps of acquiring data containing ankylosing spondylitis genetic expression from a preset database, screening the ankylosing spondylitis genetic expression data to obtain data containing a control group and a patient group at the same time, conducting differential expression gene screening on the data containing the control group and the patient group at the same time to obtain an ankylosing spondylitis abnormal expression gene dataset, conducting feature screening on the ankylosing spondylitis abnormal expression gene dataset with the recursive feature elimination method and random forest algorithm to obtain a dataset used for establishing an ankylosing spondylitis forewarning model, and training the dataset used for establishing the ankylosing spondylitis forewarning model based on the support vector machine algorithm to obtain the ankylosing spondylitis forewarning model. By the adoption of the method and device, effective warning of ankylosing spondylitis is achieved.
Description
Technical field
The present embodiments relate to biological technical field, particularly relate to a kind of ankylosing spondylitis Early-warning Model and build
Cube method and device.
Background technology
Ankylosing spondylitis (Ankylosing spondylitis, AS) be a kind of chronic, progressive self
Disease of immune system, can cause serious pain, joint stiffness and the loss of progressive spinal mobility.Should
Disease is mainly in the person between twenty and fifty in 20-30 year, and wherein, women number of patients is about the 1/10 of male sicken number,
And women state of an illness great majority are lighter.AS sickness rate in white race crowd is about 1%-3%, and China AS is ill
Rate is about 0.2%-0.6%, and wherein the patient of more than 60% is with bones of the body joint involvement, causes more than 20% AS patient
Deformity, inflammation mainly involves the bone attachment point of joint capsule, flesh key and ligament, causes local joint accretion tetanic,
Limitation of activity.At present, AS pathogenic factor is unclear, but has obvious genetic predisposition.
The general onset of AS compares concealment, in early days can be without any clinical symptoms, and some patient can show in early days
Slight General Symptoms, as weak, become thin, long-term or intermittent low grade fever, anorexia, anemia etc..Existing
For AS check method be mainly the Image Examination such as CT, x-ray inspection, HLA-B27 detection etc.
Single biological markers detection method, but Image Examination could can only be examined after pathological changes occurs in joint
Find, extend the medical diagnosis on disease cycle, make patient miss optimal treatment period, bring long-term prognosis to patient
Problem, reduces the cure rate of disease;And single biological markers detection method does not have enough sensitivity with special
The opposite sex, thus cannot be diagnosed to be exactly and be in ankylosing spondylitis in early days.
Summary of the invention
In view of this, the embodiment of the present invention provides a kind of ankylosing spondylitis Early-warning Model method for building up and device,
With overcome existing detection method cannot in time, accurately, early warning sensitively goes out ankylosing spondylitis, and then extends
In the medical diagnosis on disease cycle, make patient miss optimal treatment period, bring long-term prognosis problem, and fall to patient
The defect of low disease cured rate.
For reaching this purpose, the present invention by the following technical solutions:
First aspect, embodiments provides a kind of ankylosing spondylitis Early-warning Model method for building up, bag
Include:
The data including ankylosing spondylitis gene expression are obtained from preset data base;
The data of described ankylosing spondylitis gene expression are screened, obtain include simultaneously matched group and
The data of case group;
The described data simultaneously including matched group and case group are carried out differential gene expression screening, obtains strong
Straightforward spondylitis unconventionality expression gene data collection;
Described ankylosing spondylitis unconventionality expression gene data collection is used recursive feature elimination method and the most gloomy
Woods algorithm carries out Feature Selection, obtains the data set for building ankylosing spondylitis Early-warning Model;
Based on algorithm of support vector machine, the described data set for building ankylosing spondylitis Early-warning Model is carried out
Training, it is thus achieved that ankylosing spondylitis Early-warning Model.
In the above-mentioned methods, it is preferred that described to the described data including matched group and case group simultaneously
Carry out differential gene expression screening, before obtaining ankylosing spondylitis unconventionality expression gene data collection, also include:
The described data simultaneously including matched group and case group are carried out pretreatment.
In the above-mentioned methods, it is preferred that described based on algorithm of support vector machine to described tetanic for building
The data set of property spondylitis Early-warning Model is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, also wrap
Include:
According to crosscheck method, described ankylosing spondylitis Early-warning Model performance is tested, it is thus achieved that test
Result;
Described Early-warning Model is adjusted according to described test result.
In the above-mentioned methods, it is preferred that described based on algorithm of support vector machine to described tetanic for building
The data set of property spondylitis Early-warning Model is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, also wrap
Include:
According to described ankylosing spondylitis Early-warning Model, the patient data of pending ankylosing spondylitis early warning is entered
Row calculates and obtains early warning result;
Information is exported according to described early warning result.
In the above-mentioned methods, it is preferred that described according to described ankylosing spondylitis Early-warning Model to pending
Before the patient data of ankylosing spondylitis early warning carries out calculating acquisition early warning result, also include:
Obtaining the patient data of pending ankylosing spondylitis early warning, described patient data refers specifically to the base of patient
Because expressing data.
Second aspect, embodiments provides a kind of ankylosing spondylitis Early-warning Model and sets up device, bag
Include:
Gene expression data acquisition module, includes ankylosing spondylitis base for obtaining from preset data base
Because of the data expressed;
Garbled data acquisition module, for the data of described ankylosing spondylitis gene expression are screened,
Obtain and include matched group and the data of case group simultaneously;
Unconventionality expression gene data acquisition module, for the described number including matched group and case group simultaneously
According to carrying out differential gene expression screening, obtain ankylosing spondylitis unconventionality expression gene data collection;
Build model data acquisition module, for described ankylosing spondylitis unconventionality expression gene data centralized procurement
Carry out Feature Selection with recursive feature elimination method and random forests algorithm, obtain and be used for building rigid spine
The data set of scorching Early-warning Model;
Training module, for based on algorithm of support vector machine to described for building ankylosing spondylitis early warning mould
The data set of type is trained, it is thus achieved that ankylosing spondylitis Early-warning Model.
In said apparatus, it is preferred that also include:
Pretreatment module, for carrying out differential expression to the described data including matched group and case group simultaneously
Genescreen, obtains before ankylosing spondylitis unconventionality expression gene data collection, to described include simultaneously right
Data according to group and case group carry out pretreatment.
In said apparatus, it is preferred that also include:
Test module, for described based on algorithm of support vector machine to described pre-for building ankylosing spondylitis
The data set of alert model is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, according to crosscheck side
Described ankylosing spondylitis Early-warning Model performance is tested by method, it is thus achieved that test result;
Adjusting module, for adjusting described Early-warning Model according to described test result.
In said apparatus, it is preferred that also include:
Computing module, for described based on algorithm of support vector machine to described pre-for building ankylosing spondylitis
The data set of alert model is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, according to described tatanic
The patient data of pending ankylosing spondylitis early warning is carried out calculating acquisition early warning result by spondylitis Early-warning Model;
Output module, for exporting information according to described early warning result.
In said apparatus, it is preferred that also include:
Patient data acquisition module, for described according to described ankylosing spondylitis Early-warning Model to pending by force
Before the patient data of straightforward spondylitis early warning carries out calculating acquisition early warning result, obtain pending tatanic ridge
The patient data of post inflammation early warning, described patient data refers specifically to the gene expression data of patient.
What the technical scheme that the embodiment of the present invention provides was brought has the beneficial effect that
By the data base including ankylosing spondylitis gene expression data is screened, screening is obtained
Data acquisition with support vector machine method build ankylosing spondylitis Early-warning Model, overcome existing detection method
Cannot in time, accurately, early warning sensitively go out the defect of ankylosing spondylitis, it is achieved that to ankylosing spondylitis
Early warning, add the accuracy of early warning, improve the readability of early warning result, compensate for again single simultaneously
The limitation that biomarker sensitivity and specificity is not enough, provides supplementing and propping up of molecular level for clinical diagnosis
Hold.
Accompanying drawing explanation
Fig. 1 is the stream of a kind of ankylosing spondylitis Early-warning Model method for building up that first embodiment of the invention provides
Journey schematic diagram;
Fig. 2 is the stream of a kind of ankylosing spondylitis Early-warning Model method for building up that second embodiment of the invention provides
Journey schematic diagram;
Fig. 3 is the stream of a kind of ankylosing spondylitis Early-warning Model method for building up that third embodiment of the invention provides
Journey schematic diagram;
Fig. 4 is the knot that a kind of ankylosing spondylitis Early-warning Model that fourth embodiment of the invention provides sets up device
Structure schematic diagram;
Fig. 5 is the knot that a kind of ankylosing spondylitis Early-warning Model that fifth embodiment of the invention provides sets up device
Structure schematic diagram.
Detailed description of the invention
The present invention is described in further detail with embodiment below in conjunction with the accompanying drawings.It is understood that this
Specific embodiment described by place is used only for explaining the present invention, rather than limitation of the invention.The most also need
It is noted that for the ease of describing, accompanying drawing illustrate only part related to the present invention and not all knot
Structure.
First embodiment
Fig. 1 is the stream of a kind of ankylosing spondylitis Early-warning Model method for building up that first embodiment of the invention provides
Journey schematic diagram, the method is applicable on any device needing early warning ankylosing spondylitis.
The method of the present embodiment specifically includes:
Step 100, obtain from preset data base and include the data of ankylosing spondylitis gene expression.
Exemplary, preset data base can be comprehensive (the Gene Expression of gene expression
Omnibus, GEO) data base, it is also possible to it is any data base including ankylosing spondylitis gene expression.
Exemplary, obtain from preset data base and include that the data of ankylosing spondylitis gene expression can be adopted
Scanning for preset data base by the method for keyword search, key word can be that Chinese key is tetanic
Property spondylitis, it is also possible to be English key word Ankylosing Spondylitis, AS or
Marie-StrumpellDisease.The benefit so arranged is can to make to include ankylosing spondylitis
Chinese and English data all screen, and are used in the data more enrichment of modeling.
Exemplary, in order to improve the efficiency of search, the vector data in data base can be scanned for, as
Several ten thousand data.
Step 101, data to described ankylosing spondylitis gene expression are screened, and obtain and include simultaneously
Matched group and the data of case group.
Exemplary, the data of described ankylosing spondylitis gene expression are screened, obtains and comprise simultaneously
The data having matched group and case group can be to obtain including at least there being 30 number of cases evidences, and wherein matched group can compare
Case group is more, or matched group can be fewer than case group, or matched group is the same with than case group
Many.
Step 102, the described data simultaneously including matched group and case group are carried out difference expression gene sieve
Choosing, obtains ankylosing spondylitis unconventionality expression gene data collection.
Exemplary, the described data simultaneously including matched group and case group are carried out difference expression gene sieve
Choosing it may be that by the logarithm absolute value of the expression fold differences of case group and matched group more than or equal to 0.5, and
And the data that the P value after puppet discovery rate (False Discovery Rate, FDR) correction is less than or equal to 0.05
As ankylosing spondylitis unconventionality expression gene data.
Step 103, to described ankylosing spondylitis unconventionality expression gene data collection use recursive feature elimination method
Carry out Feature Selection with random forests algorithm, obtain the data set for building ankylosing spondylitis Early-warning Model.
Exemplary, the step of recursive feature elimination method specifically refers to first structural feature sequence coefficient, and
After be iterated, remove the feature that sequence coefficient is minimum after each iteration, finally remaining feature passed
Reduce discharging sequence.The feature of characteristic coefficient arrangement front ten can be selected for building ankylosing spondylitis Early-warning Model.
Exemplary, forest algorithm is a kind of decision Tree algorithms immediately, can be used to do and classifies, returns.Should
Algorithm is made up of multiple decision trees, compared to single decision Tree algorithms, has preferably classification and prediction effect,
Being not easy situation overfitting occur, it is the most fairly simple that it realizes process, with recursive feature elimination method phase
Carry out Feature Selection in conjunction with to ankylosing spondylitis unconventionality expression gene data collection, structure can be used in tatanic
The data of spondylitis Early-warning Model are more accurate.
Step 104, based on algorithm of support vector machine to the described number for building ankylosing spondylitis Early-warning Model
It is trained according to collection, it is thus achieved that ankylosing spondylitis Early-warning Model.
Exemplary, algorithm of support vector machine is a kind of two sorting algorithms, can be less in statistical sample amount
In the case of, it is thus achieved that good statistical law, can be used to train small sample, non-linear, high-dimensional data
Collection.
Exemplary, kernlab module can be used under R language environment to be supported vector machine modeling.
When modeling, the kernel function in kernlab module can be set to gaussian kernel function, gaussian kernel function
Width parameter σ can be 0.1, and mistake penalty factor can be 10.
In embodiments of the present invention, by the data base including ankylosing spondylitis gene expression data is entered
Row filter, the data acquisition support vector machine method obtaining screening builds ankylosing spondylitis Early-warning Model,
Overcome existing detection method cannot in time, accurately, early warning sensitively go out the defect of ankylosing spondylitis, real
Show the early warning to ankylosing spondylitis, compensate for again single biomarker sensitivity and specificity not enough simultaneously
Limitation, for clinical diagnosis provide molecular level supplement with support.
Second embodiment
Fig. 2 is the stream of a kind of ankylosing spondylitis Early-warning Model method for building up that second embodiment of the invention provides
Journey schematic diagram.The present embodiment is optimized based on above-described embodiment, in the present embodiment, described to institute
State and include the data of matched group and case group simultaneously and carry out differential gene expression screening, obtain rigid spine
Before scorching unconventionality expression gene data collection, also include: to the described number including matched group and case group simultaneously
According to carrying out pretreatment.And based on algorithm of support vector machine to described for building ankylosing spondylitis early warning mould
The data set of type is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, also include: according to intersecting inspection
Described ankylosing spondylitis Early-warning Model performance is tested by proved recipe method, it is thus achieved that test result;According to described
Test result adjusts described Early-warning Model.
Accordingly, the method for the present embodiment specifically includes:
Step 200, obtains the data including ankylosing spondylitis gene expression from preset data base.
The data of described ankylosing spondylitis gene expression are screened by step 201, obtain and include simultaneously
Matched group and the data of case group.
The described data simultaneously including matched group and case group are carried out pretreatment by step 202.
Exemplary, pre-treatment step is specially and carries on the back the data including matched group and case group successively
Scape correction process, normalized and aggregation process.
Exemplary, RMA method can be used to carry out background correction the data including matched group and case group
Process, quantile method can be used to be normalized, after normalization the data through background correction
Data medianpolish method can be used to collect.
The described data simultaneously including matched group and case group are carried out differential gene expression screening by step 203,
Obtain ankylosing spondylitis unconventionality expression gene data collection.
Step 204, uses recursive feature elimination method to described ankylosing spondylitis unconventionality expression gene data collection
Carry out Feature Selection with random forests algorithm, obtain the data set for building ankylosing spondylitis Early-warning Model.
Step 205, based on algorithm of support vector machine to the described number for building ankylosing spondylitis Early-warning Model
It is trained according to collection, it is thus achieved that ankylosing spondylitis Early-warning Model.
Step 206, tests described ankylosing spondylitis Early-warning Model performance according to crosscheck method,
Obtain test result.
Exemplary, cross validation (Cross Validation, CV) method is a kind of statistical analysis technique,
Initial data is divided into two groups by the method, wherein, one group as training set, another group as checking collection.Its
Execution process is first to be trained with training the set pair analysis model, and the mould that training obtains tested by recycling checking collection
Type.
Exemplary, the embodiment of the present invention can use 10 folding cross validation methods to ankylosing spondylitis early warning mould
Type performance is tested.Data set is divided into 10 parts, will wherein 9 numbers evidences train in turn, 1 number evidence
Verify, using the average of 10 results as the estimation to arithmetic accuracy, repeatedly 10 foldings intersections can be carried out and test
Then card averages.Advantage of this is that, test result can be made more accurate.
Step 207, adjusts described Early-warning Model according to described test result.
Exemplary, when ankylosing spondylitis Early-warning Model performance indications reach default standard, then without
Again Early-warning Model is adjusted.When the standard that ankylosing spondylitis Early-warning Model performance indications are not up to preset
Time, then adjust Early-warning Model parameter, until each performance indications of Early-warning Model all reach default standard.
In embodiments of the present invention, by the data base including ankylosing spondylitis gene expression data is entered
Row filter, the data acquisition support vector machine method obtaining screening builds ankylosing spondylitis Early-warning Model,
And use the method for cross validation constructed Early-warning Model to be tested, according to test result to early warning mould
Type is adjusted, overcome existing detection method cannot in time, accurately, early warning sensitively go out rigid spine
Scorching defect, it is achieved that the early warning to ankylosing spondylitis, adds the accuracy of early warning, makes up again simultaneously
The limitation that single biomarker sensitivity and specificity is not enough, provides the benefit of molecular level for clinical diagnosis
Fill and support.
3rd embodiment
Fig. 3 is the stream of a kind of ankylosing spondylitis Early-warning Model method for building up that third embodiment of the invention provides
Journey schematic diagram.The present embodiment is optimized based on above-described embodiment, in the present embodiment, based on support
The described data set for building ankylosing spondylitis Early-warning Model is trained by vector machine algorithm, it is thus achieved that strong
After straightforward spondylitis Early-warning Model, also include: according to described ankylosing spondylitis Early-warning Model to pending
The patient data of ankylosing spondylitis early warning carries out calculating acquisition early warning result;Export according to described early warning result
Information.And according to described ankylosing spondylitis Early-warning Model to pending ankylosing spondylitis early warning
Before patient data carries out calculating acquisition early warning result, also include: obtain pending ankylosing spondylitis early warning
Patient data, described patient data refers specifically to the gene expression data of patient.
Accordingly, the method for the present embodiment specifically includes:
Step 300, obtains the data including ankylosing spondylitis gene expression from preset data base.
The data of described ankylosing spondylitis gene expression are screened by step 301, obtain and include simultaneously
Matched group and the data of case group.
The described data simultaneously including matched group and case group are carried out differential gene expression screening by step 302,
Obtain ankylosing spondylitis unconventionality expression gene data collection.
Step 303, uses recursive feature elimination method to described ankylosing spondylitis unconventionality expression gene data collection
Carry out Feature Selection with random forests algorithm, obtain the data set for building ankylosing spondylitis Early-warning Model.
Step 304, based on algorithm of support vector machine to the described number for building ankylosing spondylitis Early-warning Model
It is trained according to collection, it is thus achieved that ankylosing spondylitis Early-warning Model.
Step 305, obtains the patient data of pending ankylosing spondylitis early warning, and described patient data refers specifically to
The gene expression data of patient.
Exemplary, the gene expression data of patient can obtain from the peripheral blood of patient.
Step 306, expresses data according to described ankylosing spondylitis Early-warning Model to the patient gene obtained and carries out
Calculate and obtain early warning result.
Exemplary, when after the patient data obtaining pending ankylosing spondylitis early warning, the number that can obtain
Calculate to Early-warning Model according to input, obtain the early warning result of warning module output.
Step 307, exports information according to described early warning result.
Exemplary, information comprises the steps that image information, text message and audio-frequency information.Wherein, figure
As information can be to treat the colored sector image that warning data is described at each age level P,
Or transverse axis represents that age, the longitudinal axis represent the colored cylindrical image of P, broken line graph etc., text envelope
Breath can be to treat the subset being likely to have AS risk in warning data to provide prompting, and audio-frequency information can
To be the sound that early warning result is described.
Exemplary, when the information of output is image information and text message, can show on a display screen
Show image information and text message;When the information of output is audio-frequency information, can be by putting equipment such as outward
Audio-frequency information play by loudspeaker.
In embodiments of the present invention, by the data base including ankylosing spondylitis gene expression data is entered
Row filter, the data acquisition support vector machine method obtaining screening builds ankylosing spondylitis Early-warning Model,
The patient gene of pending ankylosing spondylitis early warning is expressed the model that data input builds calculates early warning knot
Really, and provide result prompting, overcome existing detection method cannot in time, accurately, early warning sensitively goes out by force
The defect of straightforward spondylitis, it is achieved that the early warning to ankylosing spondylitis, improves the readability of early warning result,
Compensate for again the limitation that single biomarker sensitivity and specificity is not enough simultaneously, provide point for clinical diagnosis
Supplementing and support of sub-level.
4th embodiment
Fig. 4 is the knot that a kind of ankylosing spondylitis Early-warning Model that fourth embodiment of the invention provides sets up device
Structure schematic diagram.Device described in the present embodiment, including: gene expression data acquisition module 101, screening
Data acquisition module 102, pretreatment module 103, unconventionality expression gene data acquisition module 104, structure mould
Type data acquisition module 105, training module 106, test module 107, adjusting module 108.
Wherein, gene expression data acquisition module 101, include tatanic for acquisition from preset data base
The data of spondylitis gene expression.
Garbled data acquisition module 102, for the data of described ankylosing spondylitis gene expression are screened,
Obtain and include matched group and the data of case group simultaneously.
Pretreatment module 103, for carrying out pretreatment to the described data including matched group and case group simultaneously.
Unconventionality expression gene data acquisition module 104, for described matched group and the case group of including simultaneously
Data carry out differential gene expression screening, obtain ankylosing spondylitis unconventionality expression gene data collection.
Build model data acquisition module 105, for described ankylosing spondylitis unconventionality expression gene data collection
Use recursive feature elimination method and random forests algorithm to carry out Feature Selection, obtain and be used for building tatanic ridge
The data set of post inflammation Early-warning Model.
Training module 106, for based on algorithm of support vector machine to described for building ankylosing spondylitis early warning
The data set of model is trained, it is thus achieved that ankylosing spondylitis Early-warning Model.
Test module 107, for entering described ankylosing spondylitis Early-warning Model performance according to crosscheck method
Row test, it is thus achieved that test result.
Adjusting module 108, for adjusting described Early-warning Model according to described test result.
In embodiments of the present invention, by the data base including ankylosing spondylitis gene expression data is entered
Row filter, the data acquisition support vector machine method obtaining screening builds ankylosing spondylitis Early-warning Model,
And use the method for cross validation constructed Early-warning Model to be tested, according to test result to early warning mould
Type is adjusted, overcome existing detection method cannot in time, accurately, early warning sensitively go out rigid spine
Scorching defect, it is achieved that the early warning to ankylosing spondylitis, adds the accuracy of early warning, makes up again simultaneously
The limitation that single biomarker sensitivity and specificity is not enough, provides the benefit of molecular level for clinical diagnosis
Fill and support.
5th embodiment
Fig. 5 is the knot that a kind of ankylosing spondylitis Early-warning Model that fifth embodiment of the invention provides sets up device
Structure schematic diagram.The present embodiment is optimized based on above-described embodiment, the device described in the present embodiment,
Including: gene expression data acquisition module 101, garbled data acquisition module 102, pretreatment module 103,
Unconventionality expression gene data acquisition module 104, build model data acquisition module 105, training module 106,
Patient data acquisition module 109, computing module 110, output module 111.
Wherein, patient data acquisition module 109, for obtaining patient's number of pending ankylosing spondylitis early warning
According to, described patient data refers specifically to the gene expression data of patient.
Computing module 110, is used for according to described ankylosing spondylitis Early-warning Model pending ankylosing spondylitis
The patient data of early warning carries out calculating acquisition early warning result.
Output module 111, for exporting information according to described early warning result.
In embodiments of the present invention, by the data base including ankylosing spondylitis gene expression data is entered
Row filter, the data acquisition support vector machine method obtaining screening builds ankylosing spondylitis Early-warning Model,
The patient gene of pending ankylosing spondylitis early warning is expressed the model that data input builds calculates early warning knot
Really, and provide result prompting, overcome existing detection method cannot in time, accurately, early warning sensitively goes out by force
The defect of straightforward spondylitis, it is achieved that the early warning to ankylosing spondylitis, improves the readability of early warning result,
Compensate for again the limitation that single biomarker sensitivity and specificity is not enough simultaneously, provide point for clinical diagnosis
Supplementing and support of sub-level.
Note, above are only presently preferred embodiments of the present invention and institute's application technology principle.Those skilled in the art
It will be appreciated that the invention is not restricted to specific embodiment described here, can enter for a person skilled in the art
Row various obvious changes, readjust and substitute without departing from protection scope of the present invention.Therefore, though
So by above example, the present invention is described in further detail, but the present invention be not limited only to
Upper embodiment, without departing from the inventive concept, it is also possible to include other Equivalent embodiments more,
And the scope of the present invention is determined by scope of the appended claims.
Claims (10)
1. an ankylosing spondylitis Early-warning Model method for building up, it is characterised in that including:
The data including ankylosing spondylitis gene expression are obtained from preset data base;
The data of described ankylosing spondylitis gene expression are screened, obtain include simultaneously matched group and
The data of case group;
The described data simultaneously including matched group and case group are carried out differential gene expression screening, obtains strong
Straightforward spondylitis unconventionality expression gene data collection;
Described ankylosing spondylitis unconventionality expression gene data collection is used recursive feature elimination method and the most gloomy
Woods algorithm carries out Feature Selection, obtains the data set for building ankylosing spondylitis Early-warning Model;
Based on algorithm of support vector machine, the described data set for building ankylosing spondylitis Early-warning Model is carried out
Training, it is thus achieved that ankylosing spondylitis Early-warning Model.
Ankylosing spondylitis Early-warning Model method for building up the most according to claim 1, it is characterised in that
The described differential gene expression screening that carries out the described data simultaneously including matched group and case group, obtains strong
Before straightforward spondylitis unconventionality expression gene data collection, also include:
The described data simultaneously including matched group and case group are carried out pretreatment.
Ankylosing spondylitis Early-warning Model method for building up the most according to claim 1, it is characterised in that
Described based on algorithm of support vector machine, the described data set for building ankylosing spondylitis Early-warning Model is carried out
Training, it is thus achieved that after ankylosing spondylitis Early-warning Model, also include:
According to crosscheck method, described ankylosing spondylitis Early-warning Model performance is tested, it is thus achieved that test
Result;
Described Early-warning Model is adjusted according to described test result.
Ankylosing spondylitis Early-warning Model method for building up the most according to claim 1, it is characterised in that
Described based on algorithm of support vector machine, the described data set for building ankylosing spondylitis Early-warning Model is carried out
Training, it is thus achieved that after ankylosing spondylitis Early-warning Model, also include:
According to described ankylosing spondylitis Early-warning Model, the patient data of pending ankylosing spondylitis early warning is entered
Row calculates and obtains early warning result;
Information is exported according to described early warning result.
Ankylosing spondylitis Early-warning Model method for building up the most according to claim 4, it is characterised in that
According to described ankylosing spondylitis Early-warning Model, the patient data of pending ankylosing spondylitis early warning is counted
Calculate before obtaining early warning result, also include:
Obtaining the patient data of pending ankylosing spondylitis early warning, described patient data refers specifically to the base of patient
Because expressing data.
6. an ankylosing spondylitis Early-warning Model sets up device, it is characterised in that including:
Gene expression data acquisition module, includes ankylosing spondylitis base for obtaining from preset data base
Because of the data expressed;
Garbled data acquisition module, for the data of described ankylosing spondylitis gene expression are screened,
Obtain and include matched group and the data of case group simultaneously;
Unconventionality expression gene data acquisition module, for the described number including matched group and case group simultaneously
According to carrying out differential gene expression screening, obtain ankylosing spondylitis unconventionality expression gene data collection;
Build model data acquisition module, for described ankylosing spondylitis unconventionality expression gene data centralized procurement
Carry out Feature Selection with recursive feature elimination method and random forests algorithm, obtain and be used for building rigid spine
The data set of scorching Early-warning Model;
Training module, for based on algorithm of support vector machine to described for building ankylosing spondylitis early warning mould
The data set of type is trained, it is thus achieved that ankylosing spondylitis Early-warning Model.
Ankylosing spondylitis Early-warning Model the most according to claim 6 sets up device, it is characterised in that
Also include:
Pretreatment module, for carrying out differential expression to the described data including matched group and case group simultaneously
Genescreen, obtains before ankylosing spondylitis unconventionality expression gene data collection, to described include simultaneously right
Data according to group and case group carry out pretreatment.
Ankylosing spondylitis Early-warning Model the most according to claim 6 sets up device, it is characterised in that
Also include:
Test module, for described based on algorithm of support vector machine to described pre-for building ankylosing spondylitis
The data set of alert model is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, according to crosscheck side
Described ankylosing spondylitis Early-warning Model performance is tested by method, it is thus achieved that test result;
Adjusting module, for adjusting described Early-warning Model according to described test result.
Ankylosing spondylitis Early-warning Model the most according to claim 6 sets up device, it is characterised in that
Also include:
Computing module, for described based on algorithm of support vector machine to described pre-for building ankylosing spondylitis
The data set of alert model is trained, it is thus achieved that after ankylosing spondylitis Early-warning Model, according to described tatanic
The patient data of pending ankylosing spondylitis early warning is carried out calculating acquisition early warning result by spondylitis Early-warning Model;
Output module, for exporting information according to described early warning result.
Ankylosing spondylitis Early-warning Model the most according to claim 9 sets up device, it is characterised in that
Also include:
Patient data acquisition module, for described according to described ankylosing spondylitis Early-warning Model to pending by force
Before the patient data of straightforward spondylitis early warning carries out calculating acquisition early warning result, obtain pending tatanic ridge
The patient data of post inflammation early warning, described patient data refers specifically to the gene expression data of patient.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610516917.6A CN105938521A (en) | 2016-07-04 | 2016-07-04 | Ankylosing spondylitis forewarning model establishing method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610516917.6A CN105938521A (en) | 2016-07-04 | 2016-07-04 | Ankylosing spondylitis forewarning model establishing method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105938521A true CN105938521A (en) | 2016-09-14 |
Family
ID=56873493
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610516917.6A Pending CN105938521A (en) | 2016-07-04 | 2016-07-04 | Ankylosing spondylitis forewarning model establishing method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105938521A (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101506379A (en) * | 2006-05-22 | 2009-08-12 | 临床基因组学有限公司 | Detection method |
CN103502473A (en) * | 2011-03-01 | 2014-01-08 | 耶鲁大学 | Predicting gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) |
-
2016
- 2016-07-04 CN CN201610516917.6A patent/CN105938521A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101506379A (en) * | 2006-05-22 | 2009-08-12 | 临床基因组学有限公司 | Detection method |
CN103502473A (en) * | 2011-03-01 | 2014-01-08 | 耶鲁大学 | Predicting gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) |
Non-Patent Citations (4)
Title |
---|
张力 等: "基于随机森林和特征选择方法的蛋白质热稳定性影响因素预测", 《现代食品科技》 * |
李小波: "多步骤降维的肿瘤特征基因选择方法", 《复旦学报(自然科学版)》 * |
李贞双 等: "基于支持向量机的结肠癌分类方法研究", 《电能知识与技术》 * |
高山 等: "《R语一言与Bioconductor生物信息学应用》", 31 January 2014 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107066791A (en) | A kind of aided disease diagnosis method based on patient's assay | |
CN107785057B (en) | Medical data processing method, device, storage medium and computer equipment | |
CN110051324B (en) | Method and system for predicting death rate of acute respiratory distress syndrome | |
KR20170061222A (en) | The method for prediction health data value through generation of health data pattern and the apparatus thereof | |
CN108511056A (en) | Therapeutic scheme based on patients with cerebral apoplexy similarity analysis recommends method and system | |
CN110503635B (en) | Hand bone X-ray film bone age assessment method based on heterogeneous data fusion network | |
CN110097975A (en) | A kind of nosocomial infection intelligent diagnosing method and system based on multi-model fusion | |
CN111584021A (en) | Medical record information verification method and device, electronic equipment and storage medium | |
CN109065171A (en) | The construction method and system of Kawasaki disease risk evaluation model based on integrated study | |
CN106682412A (en) | Diabetes prediction method based on medical examination data | |
CN109480864A (en) | A kind of schizophrenia automatic evaluation system based on nervous functional defects and machine learning | |
CN106529110A (en) | Classification method and equipment of user data | |
CN109273094A (en) | A kind of construction method and building system of the Kawasaki disease risk evaluation model based on Boosting algorithm | |
CN112541066A (en) | Text-structured-based medical and technical report detection method and related equipment | |
CN110874409A (en) | Disease grading prediction system, method, electronic device and readable storage medium | |
CN112052874B (en) | Physiological data classification method and system based on generation countermeasure network | |
CN108595432B (en) | Medical document error correction method | |
CN104361245B (en) | Measurement data-processing system and method | |
CN111524570A (en) | Ultrasonic follow-up patient screening method based on machine learning | |
CN114662477A (en) | Stop word list generating method and device based on traditional Chinese medicine conversation and storage medium | |
CN107122582A (en) | Towards the diagnosis and treatment class entity recognition method and device of multi-data source | |
CN106951710A (en) | CAP data systems and method based on privilege information Learning support vector machine | |
CN113539412A (en) | Chinese herbal medicine recommendation system based on deep learning | |
CN108831560B (en) | Method and device for determining medical data attribute data | |
CN114582511B (en) | Bronchiectasis acute exacerbation early warning method, device, equipment and medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160914 |