CN109817337A - A kind of appraisal procedure and similar disorder differentiating method of single disease sample Pathway Activation degree - Google Patents
A kind of appraisal procedure and similar disorder differentiating method of single disease sample Pathway Activation degree Download PDFInfo
- Publication number
- CN109817337A CN109817337A CN201910091441.XA CN201910091441A CN109817337A CN 109817337 A CN109817337 A CN 109817337A CN 201910091441 A CN201910091441 A CN 201910091441A CN 109817337 A CN109817337 A CN 109817337A
- Authority
- CN
- China
- Prior art keywords
- gene
- sample
- access
- disease sample
- full
- 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
Abstract
The invention discloses the appraisal procedures and similar disorder differentiating method of a kind of single disease sample Pathway Activation degree；Full-mesh network is constructed to every access, and is connected original in access while important as its, the company of addition is while as its background；Using gene present in access as important gene, other genes are as background genes；To every company side in full-mesh network, the difference value of disease sample and normal sample is calculated, and calculates difference value conspicuousness；Calculate the fold differences of each gene expression value in disease sample and normal sample；Calculate the enrichment degree of important node and Lian Bian in the node of each full-mesh network and Lian Bian ranking, the activity as respective channels.The differentiation similar disorder characterized by the activity of access.The present invention can effectively calculate the activity of every access in single disease sample, and the gene expression matrix of disease sample higher-dimension, small sample is converted into the expression matrix of Pathway Activation degree, and for distinguishing similar disorder, accuracy is high.
Description
Technical field
The present invention relates to field of bioinformatics, are related to the appraisal procedure and phase of a kind of single disease sample Pathway Activation degree
Like disease differentiating method.
Background technique
Studies have shown that gene and gene product be not individually play a role, but participate in it is complicated, mutually close
Collaboration plays a role in the network of connection.Common includes access, gene transcription regulation net with biological structure existing for latticed form
Network, protein-protein interaction network, wherein access can with the biological processes in reacting cells, as biological metabolism, signal transmitting and
Growth cycle, in conjunction with the effective biological information of access data mining for from the molecular mechanism of functional perspective announcement organism to pass
It is important.
Imbalance of the occurrence and development of disease usually with important access is closely related, and identifies the access of these imbalances and quantifies it
Imbalance degree is significant to the research of disease.
Pathway Activation degree (pathway activity) can be used for measuring the imbalance degree of access.In addition, although similar multiple
The clinical symptoms of miscellaneous disease are similar, but the mechanism of various disease occurrence and development is different, therefore the state of activation of access can be used as
Distinguish the index of similar disorder.Current existing some models and method are used to assess the activation of access in disease generating process
Degree, they are different for the definition of Pathway Activation degree and calculation method, such as Han[1]It is proposed the method benefit of entitled PROPS
Pathway Activation degree is calculated with Gauss Bayesian network.Young and Craft[2]Provide three kinds of Pathway Activation degree calculation methods:
PCA, NTC and GED.PCA using Principal Component Analysis extract the gene expression data based on every access in principal component as
The activity of access；NTC method is the Europe between the gene expression data calculating disease sample based on every access and normal sample
Activity of the formula distance as access；GED is to poor in the gene expression data of every access in normal sample and disease sample
Allochoric gene marking, according to gene marking value passage activity feature.And single disease sample is considered from access angle
This specific state is most important for the molecular mechanism for disclosing complex disease from system level, but current model and side
Method does not consider the specific state of single disease sample from access angle.
In addition, although existing some models and method can distinguish similar disorder, such as Winter[3]Propose one kind
The method NetRank for improving Page sequence extracts sequence according to the ranking of the neighbor node of gene in a network to gene order
Forward gene is as the feature for distinguishing similar disorder.Cun andPropose the feature choosing based on support vector machines
Selection method stSVM extracts efficient gene marker as the feature for distinguishing similar disorder.Zhang etc.[5]Propose a use
In the gene that the frame CNS of abstraction function feature, this method are enriched with using mobile equilibrium model aggregation by identical function, thus
To the functional module that can distinguish two kinds of similar disorders to greatest extent, these functional modules are extracted as the spy for distinguishing similar disorder
Sign.But the classification accuracy that the feature extracted based on these methods carries out similar disorder classification needs to be further increased.
A kind of single disease sample Pathway Activation degree is assessed and effective district divides the side of similar disorder therefore, it is necessary to provide
Method.
[1]Han,L.et al.A probabilistic pathway score(PROPS)for classification
with applications to inflammatory bowel disease.Bioinformatics,2017；34(6):
985-993.
[2]Young,M.R.and Craft,D.L.Pathway-informed classification system
(PICS)for cancer analysis using gene expression data.Cancer informatics,2016；
15:151-161.
[3]Winter C,Kristiansen G,Kersting S,et al.Google goes cancer:
improving outcome prediction for cancer patients by network-based ranking of
marker genes[J].PLoS computational biology,2012,8(5):e1002511.
[4]Cun Y,H.Network and data integration for biomarker
signature discovery via network smoothed t-statistics[J].PloS one,2013,8(9):
e73074.
[5]Zhang C,Liu J,Shi Q,et al.Comparative network stratification
analysis for identifying functional interpretable network biomarkers[J].BMC
bioinformatics,2017,18(3):48.
Summary of the invention
The technical problem to be solved by the present invention is in view of the shortcomings of the prior art, provide a kind of single disease sample access
The appraisal procedure and similar disorder differentiating method of activity, it is available can effective district divide feature-disease sample of similar disorder
This Pathway Activation degree carries out similar disorder classification based on this feature, and classification accuracy is high.
In order to solve the above technical problems, the technical scheme adopted by the invention is that:
A kind of appraisal procedure of single disease sample Pathway Activation degree, the activity of every access include even side activity and
Gene activation degree, for every access in disease sample, activity appraisal procedure the following steps are included:
Step 1, for all genes in the access, if not connecting side between two genes, the company of addition side leads to this
Road is built into full-mesh network (i.e. node has the network for connecting side between any two)；
Connect original in the access while important as its, the company of addition is while as its background；
Using gene present in the access as important gene, the gene being not present in the access (is present in the disease sample
The gene of other accesses in this) it is used as background genes；
Step 2, to every company side in full-mesh network, be primarily based on n normal sample, calculate its connection two bases
Because of the Pearson correlation coefficients of the expression value in this n sample, it is denoted as PCCn；Single disease is added in n normal sample again
Sick sample calculates the Pearson correlation coefficients of expression value of two genes of its connection in this n+1 sample, is denoted as PCCn+1；
Pass through PCCn+1With PCCnΔ PCC is obtained as differencen, connect difference value of the side in the disease sample and normal sample as this；And it comments
Estimate the conspicuousness of difference value；
To each gene in full-mesh network, it is calculated in the difference of the disease sample and expression value in n normal sample
Different multiple；
Even sides all in full-mesh network are ranked up by step 4 according to the conspicuousness size of difference value, by full-mesh net
All genes are ranked up according to fold differences size in network；
Step 5, according to ranking results, calculate company in full-mesh network it is important in/gene order while/gene (positive mark
Label) enrichment degree, as connecting side/gene activity in respective channels.
Further, in the step 2, expression value of two genes of certain company side connection in n sample is calculated
Pearson correlation coefficients PCCnFormula are as follows:
Wherein, x1And x2Respectively indicate expression value of two genes of this company side connection in n sample, covn(x1,
x2) indicate x1And x2Covariance,WithRespectively indicate x1And x2Standard deviation.
Further, in the step 2, the conspicuousness based on Z test (z-test) assessment difference value:
Wherein, Z value indicates Δ PCCnConspicuousness.
Further, in the step 3, for any gene, it is calculated in the disease sample and table in n normal sample
Up to the formula of the fold differences FC of value are as follows:
Wherein, b indicates expression value of the gene in the disease sample,Indicate the gene in n normal sample
The mean value of expression value.
Further, it in the step 5, to any access, is calculated by the following formula it and connects side/node activity:
Wherein, I indicates all important set that even side/gene is constituted, rank in full-mesh networkiIt indicates to press in step 4
When according to ascending order arrangement, i-th company side/gene sequence in I, M indicates important even side/gene sum, N table in full-mesh network
Show that background connects side/gene sets sum；The formula is using AUC from even side/gene angle calculation in the company of full-mesh network
In/gene order it is important even while/gene (positive label) enrichment degree, the activity as access.
A kind of similar disorder differentiating method, comprising the following steps:
Firstly, calculating the Pathway Activation degree of each disease sample, and the aisled even side of single disease sample institute is activated
Degree and gene activation degree connect into a vector, the feature vector as the disease sample；In all disease sample feature vectors
It is identical with the corresponding feature of dimension, that is, correspond to the company's side activity or gene activation degree of same access；
It then, is input with the feature vector of known disease sample, the tag along sort of each known disease sample is output,
Training classifier；
Finally, the feature vector of unidentified illness sample is inputted trained classifier, its tag along sort is obtained.
Further, the classifier is random forest grader.
The utility model has the advantages that
The present invention can effectively calculate the activity of every access in single disease sample, by disease sample higher-dimension, small sample
Gene expression matrix be converted into the expression matrix of Pathway Activation degree, solve and do not consider single disease in other feature extracting methods
The problem of the specificity of sick sample.The activity of calculated access can be used for distinguishing similar disorder, and accuracy is high.
Detailed description of the invention
Fig. 1 is the frame diagram of (PASS) of the invention；
Fig. 2 be of the invention (PASS) and NetRank, stSVM, CNS, PCA, NTC, GED, PROPS method ROC curve and its
Under area (AUC) comparison figure；
Fig. 3 is the significant difference of access in the two kinds of similar disorder samples obtained based on the Pathway Activation degree of the invention extracted
Property analysis.
Fig. 4 is that the significant difference that the Pathway Activation degree extracted based on the present invention is obtained expresses known disease gene in access
Enriching analysis.
Specific embodiment
As shown in Figure 1, the present invention provides a kind of appraisal procedure of single disease sample Pathway Activation degree, every access
Activity includes connecting side activity and gene activation degree, for every access in disease sample, activity appraisal procedure packet
Include following steps:
One, the pretreatment of access data
For all genes in access, if not connecting side between two genes, which is built by the company of addition side
Full-mesh network (i.e. node has the network for connecting side between any two)；
Connect original in the access while important as its, the company of addition is while as its background；
Using gene present in the access as important gene, the gene being not present in the access (is present in the disease sample
The gene of other accesses in this) it is used as background genes；
Two, the significance of difference on side is calculated
To every company side in full-mesh network, it is primarily based on n normal sample, calculates two genes of its connection at this
The Pearson correlation coefficients of expression value in n sample, are denoted as PCCn；Single disease sample is added in n normal sample again,
The Pearson correlation coefficients for calculating expression value of two genes of its connection in this n+1 sample, are denoted as PCCn+1；Pass through
PCCn+1With PCCnΔ PCC is obtained as differencen, connect difference value of the side in the disease sample and normal sample as this；And assess difference
Different value Δ PCCnConspicuousness；
Calculate the Pearson correlation coefficients PCC of expression value of two genes of certain company side connection in n samplenPublic affairs
Formula are as follows:
Wherein, x1And x2Respectively indicate expression value of two genes of this company side connection in n sample, covn(x1,
x2) indicate x1And x2Covariance,WithRespectively indicate x1And x2Standard deviation；
ΔPCCnConspicuousness pass through z-test assess:
Three, the significance of difference of calculate node
The expression formula of each gene fold differences of expression value in single disease sample and normal sample are as follows:
Wherein, b indicates expression value of the gene in the disease sample,Indicate the gene in n normal sample
The mean value of expression value.
Four, Pathway Activation degree is assessed
The activity of access is calculated by the following formula to obtain:
Wherein, I indicates all important set that even side/gene is constituted, rank in full-mesh networkiIndicate i-th in I
Side/gene indicates in full-mesh network according to the position after conspicuousness size/fold differences size ascending sort of difference value, M
Important even side/gene sum, N indicate that background connects side/gene sets sum；The formula is using AUC respectively from Lian Bian and base
Because (node) angle calculation each full-mesh network company while and gene order in it is important even while/gene enrichment degree,
Activity as access.
Access can be calculated in the single disease sample based on the Pathway Activation degree that single disease sample is assessed
Activity situation, solve the problems, such as not consider in other feature extracting methods each disease sample specificity.
The present invention also provides a kind of similar disorder differentiating methods, comprising the following steps:
Firstly, calculating the Pathway Activation degree of each disease sample, and the aisled even side of single disease sample institute is activated
Degree and gene activation degree connect into a vector, the feature vector as the disease sample；
It then, is input with the feature vector of known disease sample, the tag along sort of each known disease sample is output,
Training classifier；
Finally, the feature vector of unidentified illness sample is inputted trained classifier, its tag along sort is obtained.
Random forest grader can be used in the classifier.
Five, experimental verification
In order to verify the validity of this method, based on two kinds of similar disorders in inflammatory bowel disease --- it regional enteritis and bursts
Four data sets of ulcer enteritis are verified.Four data sets of regional enteritis and ulcerative colitis derive from GEO data
Library (https: //www.ncbi.nlm.nih.gov/geo/), respectively GSE9686, GSE3365, GSE36807,
GSE71730 includes 61 ulcerative colitis samples and 105 regional enteritis samples altogether.Whole mankind's access data from
KEGG database (https: //www.kegg.jp/), shares 294 accesses.
In order to evaluate the accuracy and function interpretation of this method classification, following three analyses of progress:
(1) accuracy of analysis classification
All samples that the part is concentrated for four data are analyzed together.For of the invention (PASS) with
Each method in NetRank, stSVM, CNS, PCA, NTC, GED, PROPS, be based respectively on its extraction feature, building with
Machine forest classified device, and three folding cross validation methods are applied, sample set is divided into 3 subsets, each subset is done once respectively
Verifying collection, remaining 2 subset obtain 3 classifiers as training set, the sample concentrated using classifier to corresponding verifying
Classify, obtains classification results；Repeat 500 three folding cross validations (different demarcation is carried out to sample set every time), base
True positive rate (TPR) and false positive rate (FPR) are calculated in all classification results, draws ROC curve.Using ROC and AUC index
Classification of assessment result.AUC value be ROC curve line under area, ROC and AUC experimental result is as shown in Fig. 2, can from Fig. 2
Out, AUC value of the invention is superior to other methods.
(2) significance of difference of the access in two kinds of similar disorder samples is analyzed
The sample that the part is concentrated for four data is analyzed respectively.It is true using the t method of inspection to each access
Whether fixed difference of its activity in two kinds of similar disorder samples of each data set is significant.Step are as follows: be first based respectively on this hair
Bright method calculates activity of the access in each disease sample, then using the t value calculation formula computational representation access
The t value of difference degree of the activity in two kinds of similar disorder samples, then t dividing value table is looked into, determine slogan banner in t dividing value table (freely
Degree) the sum of number of two kinds of disease samples -2 in=data set, it is worth the corresponding vertical mark P of cell for t, if P≤0.05, explanation
Significant difference of the activity of the access in both similar disorder samples.The corresponding P value of all accesses is counted, such as Fig. 3 institute
Show, from figure 3, it can be seen that the corresponding P value of most of accesses is less than or equal to 0.05, illustrate the activities of most of accesses this two
Significant difference in kind similar disorder sample.
Enrichment degree of the known disease gene in differential expression access in (3) two kinds of similar disorders.
The sample that the part is concentrated for four data is analyzed respectively.It is less than or equal to corresponding P value is obtained in (2)
0.05 access distinguishes enrichment journey of the known disease gene in these accesses in two kinds of similar disorders as differential expression access
Degree.
The P value for being calculated known disease gene enrichment degree in differential expression access is examined by hypergeometry:
Wherein, N is the gene number in all accesses, and M is the quantity of known disease gene, and n is in differential expression access
Gene number, m are the quantity of the known disease gene in difference access.P value is smaller, illustrates known disease gene in differential expression
Enrichment degree in access is higher.- the log obtained based on four data sets10The result of P is as shown in figure 4, can from Fig. 4
Out ,-log10P is all larger than equal to 1.3, i.e., P value is less than or equal to 0.05, illustrates known richness of the disease gene in differential expression access
Collection degree is very high.
Fig. 3 and Fig. 4's the result shows that, the Pathway Activation degree for the single disease sample that the method for the present invention extracts can be effective
The difference between similar disorder is embodied, the Pathway Activation degree calculation method provided through the invention can effectively distinguish both phases
Like disease.
The experimental results showed that the method for the present invention has good classification accuracy and stability.
Claims (7)
1. a kind of appraisal procedure of single disease sample Pathway Activation degree, which is characterized in that the activity of every access includes connecting
Side activity and gene activation degree, for every access in disease sample, activity appraisal procedure the following steps are included:
Step 1, for all genes in the access, if not connecting side between two genes, the company of addition side, by the access structure
Build up full-mesh network；
Connect original in the access while important as its, the company of addition is while as its background；
Using gene present in the access as important gene, the gene being not present in the access is as background genes；
Step 2, to every company side in full-mesh network, be primarily based on n normal sample, calculate its connection two genes exist
The Pearson correlation coefficients of expression value in this n sample, are denoted as PCCn；Single disease sample is added in n normal sample again
This, calculates the Pearson correlation coefficients of expression value of two genes of its connection in this n+1 sample, is denoted as PCCn+1；Pass through
PCCn+1With PCCnΔ PCC is obtained as differencen, connect difference value of the side in the disease sample and normal sample as this；And assess difference
The conspicuousness of different value；
To each gene in full-mesh network, it is calculated in the difference times of the disease sample and expression value in n normal sample
Number；
Even sides all in full-mesh network are ranked up by step 4 according to the conspicuousness size of difference value, will be in full-mesh network
All genes are ranked up according to fold differences size；
Step 5, according to ranking results, calculate company in full-mesh network it is important in/gene order while/gene enrichment journey
Degree, as connecting side/gene activity in respective channels.
2. the appraisal procedure of single disease sample Pathway Activation degree according to claim 1, which is characterized in that the step
In 2, the Pearson correlation coefficients PCC of expression value of two genes of certain company side connection in n sample is calculatednFormula
Are as follows:
Wherein, x1And x2Respectively indicate expression value of two genes of this company side connection in n sample, covn(x1,x2) table
Show x1And x2Covariance,WithRespectively indicate x1And x2Standard deviation.
3. the appraisal procedure of single disease sample Pathway Activation degree according to claim 1, which is characterized in that the step
In 2, the conspicuousness based on Z test assessment difference value:
Wherein, Z value indicates Δ PCCnConspicuousness.
4. the appraisal procedure of single disease sample Pathway Activation degree according to claim 1, which is characterized in that the step
In 3, for any gene, it is calculated in the formula of the disease sample and the fold differences FC of expression value in n normal sample are as follows:
Wherein, b indicates expression value of the gene in the disease sample,Indicate expression value of the gene in n normal sample
Mean value.
5. the appraisal procedure of single disease sample Pathway Activation degree according to claim 1, which is characterized in that the step
In 5, to any access, it is calculated by the following formula it and connects side/node activity:
Wherein, I indicates all important set that even side/gene is constituted, rank in full-mesh networkiIt indicates in step 4 according to ascending order
When arrangement, i-th company side/gene sequence in I, M indicates that important even side/gene sum, N indicate background in full-mesh network
Even side/gene sets sum；The formula using AUC from even in the angle calculation of/gene in the company of full-mesh network/gene
Important even side/gene enrichment degree, the activity as access in sequence.
6. a kind of similar disorder differentiating method, which comprises the following steps:
Firstly, calculate the Pathway Activation degree of each disease sample, and by the aisled even side activity of single disease sample institute and
Gene activation degree connects into a vector, the feature vector as the disease sample；
It then, is input with the feature vector of known disease sample, the tag along sort of each known disease sample is output, training
Classifier；
Finally, the feature vector of unidentified illness sample is inputted trained classifier, its tag along sort is obtained.
7. similar disorder differentiating method according to claim 6, which is characterized in that the classifier is random forest classification
Device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910091441.XA CN109817337B (en) | 2019-01-30 | 2019-01-30 | Method for evaluating channel activation degree of single disease sample and method for distinguishing similar diseases |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910091441.XA CN109817337B (en) | 2019-01-30 | 2019-01-30 | Method for evaluating channel activation degree of single disease sample and method for distinguishing similar diseases |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109817337A true CN109817337A (en) | 2019-05-28 |
| CN109817337B CN109817337B (en) | 2020-09-08 |
Family
ID=66605807
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910091441.XA Active CN109817337B (en) | 2019-01-30 | 2019-01-30 | Method for evaluating channel activation degree of single disease sample and method for distinguishing similar diseases |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109817337B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093119A (en) * | 2013-01-24 | 2013-05-08 | 南京大学 | Method for recognizing significant biologic pathway through utilization of network structural information |
| CN103180462A (en) * | 2010-10-06 | 2013-06-26 | 拜奥默里克斯公司 | Method for determining biological pathway activity |
| CN104781423A (en) * | 2012-09-18 | 2015-07-15 | 新加坡科技研究局 | Grouping for classifying gastric cancer |
| US20150302140A1 (en) * | 2012-11-02 | 2015-10-22 | H. Lee Moffit Cancer Center And Research Institute, Inc. | In silico identification of cancer molecular signaling pathways and drug candidates |
| US20160026752A1 (en) * | 2014-05-30 | 2016-01-28 | Five3 Genomics, Llc | Systems And Methods For Comprehensive Analysis Of Molecular Profiles Across Multiple Tumor And Germline Exomes |
| CN107004067A (en) * | 2015-03-23 | 2017-08-01 | 国际商业机器公司 | The relevance evaluation of biological pathway and visualization |
| CN107133492A (en) * | 2017-05-02 | 2017-09-05 | 温州大学 | A kind of method that gene pathway is recognized based on PAGIS |
| CN107203704A (en) * | 2017-05-02 | 2017-09-26 | 温州大学 | A kind of method that gene pathway is recognized based on GSA |
| CN107220526A (en) * | 2017-05-02 | 2017-09-29 | 温州大学 | A kind of method that gene pathway is recognized based on PADOG |
| CN108763862A (en) * | 2018-05-04 | 2018-11-06 | 温州大学 | A kind of active method of derivation gene pathway |
| CN108763864A (en) * | 2018-05-04 | 2018-11-06 | 温州大学 | A method of evaluation biological pathway sample state |
-
2019
- 2019-01-30 CN CN201910091441.XA patent/CN109817337B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103180462A (en) * | 2010-10-06 | 2013-06-26 | 拜奥默里克斯公司 | Method for determining biological pathway activity |
| CN104781423A (en) * | 2012-09-18 | 2015-07-15 | 新加坡科技研究局 | Grouping for classifying gastric cancer |
| US20150302140A1 (en) * | 2012-11-02 | 2015-10-22 | H. Lee Moffit Cancer Center And Research Institute, Inc. | In silico identification of cancer molecular signaling pathways and drug candidates |
| CN103093119A (en) * | 2013-01-24 | 2013-05-08 | 南京大学 | Method for recognizing significant biologic pathway through utilization of network structural information |
| US20160026752A1 (en) * | 2014-05-30 | 2016-01-28 | Five3 Genomics, Llc | Systems And Methods For Comprehensive Analysis Of Molecular Profiles Across Multiple Tumor And Germline Exomes |
| CN107004067A (en) * | 2015-03-23 | 2017-08-01 | 国际商业机器公司 | The relevance evaluation of biological pathway and visualization |
| CN107133492A (en) * | 2017-05-02 | 2017-09-05 | 温州大学 | A kind of method that gene pathway is recognized based on PAGIS |
| CN107203704A (en) * | 2017-05-02 | 2017-09-26 | 温州大学 | A kind of method that gene pathway is recognized based on GSA |
| CN107220526A (en) * | 2017-05-02 | 2017-09-29 | 温州大学 | A kind of method that gene pathway is recognized based on PADOG |
| CN108763862A (en) * | 2018-05-04 | 2018-11-06 | 温州大学 | A kind of active method of derivation gene pathway |
| CN108763864A (en) * | 2018-05-04 | 2018-11-06 | 温州大学 | A method of evaluation biological pathway sample state |
Non-Patent Citations (1)
| Title |
|---|
| PITAK SOOTANAN: ""Pathway Activity Inferences with Negatively Correlated Features for Pancreatic Cancer Classification"", 《INTERNATIONAL CONFERENCE ON BIOMEDICAL ENGINEERING & INFORMATICS》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109817337B (en) | 2020-09-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | A classification of public transit users with smart card data based on time series distance metrics and a hierarchical clustering method | |
| Pyne et al. | Automated high-dimensional flow cytometric data analysis | |
| Lee et al. | Discovering genetic ancestry using spectral graph theory | |
| CN106709272B (en) | Method and system based on decision template prediction drug target protein interaction relationship | |
| CN105260437B (en) | Text classification feature selection approach and its application in biological medicine text classification | |
| Zhang et al. | University of Illinois at Urbana–Champaign | |
| CN102262642B (en) | Web image search engine and realizing method thereof | |
| CN106203483B (en) | A kind of zero sample image classification method based on semantic related multi-modal mapping method | |
| CN106021984A (en) | Whole-exome sequencing data analysis system | |
| CN106339416A (en) | Grid-based data clustering method for fast researching density peaks | |
| CN110459264A (en) | Based on grad enhancement decision tree prediction circular rna and disease associated method | |
| Chen et al. | A gene profiling deconvolution approach to estimating immune cell composition from complex tissues | |
| van der Ploeg et al. | Feature selection and validated predictive performance in the domain of Legionella pneumophila: a comparative study | |
| Liu et al. | Why can deep convolutional neural networks improve protein fold recognition? A visual explanation by interpretation | |
| CN109856307B (en) | Metabolic component molecular variable comprehensive screening technology | |
| CN103473416A (en) | Protein-protein interaction model building method and device | |
| CN111863135B (en) | False positive structure variation filtering method, storage medium and computing device | |
| CN110010204A (en) | Prognosis biomarker recognition methods based on converged network and more marking strategies | |
| CN103310126A (en) | Classification-model building method and device | |
| Ji et al. | Analysis of music/speech via integration of audio content and functional brain response | |
| CN109817337A (en) | A kind of appraisal procedure and similar disorder differentiating method of single disease sample Pathway Activation degree | |
| Maryoosh et al. | A Review: Data Mining Techniques and Its Applications | |
| CN109903814A (en) | The interactive AI artificial intelligence gene alaysis system that knowledge based classification obtains | |
| CN111710363B (en) | Method and device for determining metabolite pairing relation | |
| Walsh et al. | Ab initio and homology based prediction of protein domains by recursive neural networks |
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 |