CN106650241A - Prediction and diagnosis method for complex disease based on Multi-Omics - Google Patents

Abstract

The invention relates to a prediction and diagnosis method for complex disease based on Multi-Omics, which comprises acquiring the first genome message of subjects, matching the first genome message with the relationship message between acquired the genome mutation sites and complex disease, calculating corresponding discrimination CDDS value, comparing the relationship of size between the CDDS value and pre-stored CDDS value and getting the result. According to the comparison results, we can analyze the probability of the complex disease tested by the subjects, which can predict and diagnose early whether the subjects suffer from the tested complex disease, and further improve the health level and promote the process of the prevention and cure.

Description

A kind of Complex Diseases prediction and diagnostic method based on multigroup

Technical field

The present invention relates to medical domain, more particularly to a kind of Complex Diseases prediction and diagnostic method based on multigroup.

Background technology

Human diseases are divided into two big class, Mendelian inheritance disease and non-Mendelian heredity according to the relation of heredity Disease, i.e.,：Simplicity disease and Complex Diseases.The former is more to belong to rare disease, and majority can find clearly heredity phase Disease-causing gene is closed, and incidence meets mendelian inheritance, can be according to classical gene genetic rule to disease Accurately predicted and diagnosed.And the latter is more belongs to common disease, without obvious Mendelian inheritance phenomenon, but There is certain genetic predisposition, often there is more diseases predisposing gene, it is impossible to by classical gene genetic rule to disease Disease is predicted and diagnoses, such as：Tumour, diabetes, tuberculosis etc..The application patent is mainly for Complex Diseases Prediction and diagnosis propose feasible technical solution.

The content of the invention

The technical problem to be solved is for the deficiencies in the prior art, there is provided a kind of complexity based on multigroup Property disease forecasting and diagnostic method.

The technical scheme that the present invention solves above-mentioned technical problem is as follows：A kind of Complex Diseases prediction based on multigroup and Diagnostic method, it is characterised in that comprise the following steps：

Step 1, the first genomic information for obtaining subject；

Step 2, associating first genomic information and acquired genome mutation site and Complex Diseases Property information matched, calculate the corresponding Complex Diseases of first genomic information and differentiate scoring CDDS values；

Step 3, relatively more described CDDS values and the magnitude relationship of the CDDS a reference values for prestoring, obtain comparative result；

Step 4, according to the comparative result, determine that the subject suffers from the probability of institute's detection of complex disease.

The invention has the beneficial effects as follows：By by the genomic information of subject and acquired genome mutation site with The relationship information of Complex Diseases is matched, and calculates the corresponding CDDS values of the genomic information, and compare CDDS values with it is pre- The magnitude relationship of the CDDS a reference values deposited determines that subject suffers from the probability of institute's detection of complex disease, such that it is able to be to subject It is no to be prevented and early diagnosed with institute's detection of complex disease, the general level of the health of crowd is improved, advance from curing the disease to preventing The process of disease.

On the basis of above-mentioned technical proposal, the present invention can also do following improvement.

Further, step 4 includes：When the comparative result is that the CDDS values are more than the CDDS a reference values, really The fixed subject suffers from the probability of institute's detection of complex disease more than 85%；Or,

When the comparative result is that the CDDS values are less than the CDDS a reference values, determines that the subject suffers from and detected The probability of Complex Diseases is less than 15%.

Further, before step 1, also include：

Step 5, according to obtain healthy population genomic information and target disease crowd's genomic information, obtain genome Variant sites and Complex Diseases relationship information.

Further, step 5 includes：

Step 5.1, the acquisition healthy population genomic information and the target disease crowd genomic information；

Step 5.2, the corresponding multiple models of target disease are built based on multigroup learning；

Step 5.3, respectively by the healthy population genomic information and the target disease crowd genomic information and institute State multiple models to be mapped, obtain healthy population model set and target disease crowd's base model set；

Step 5.4, for each model in the plurality of module, the healthy population model set and institute are calculated respectively The corresponding CDDS values of the second genomic information for belonging to same model in target disease crowd's base model set are stated, multiple institutes are obtained State the second genomic information and distinguish corresponding multiple CDDS values；

Step 5.5, multiple second genomic informations and multiple CDDS values are stored correspondingly, form base Because of group variant sites and the relationship information of Complex Diseases, the relationship information is the genomic information and CDDS values Corresponding relation.

Further, first genomic information, the healthy population genomic information and the target disease crowd Genomic information is obtained by chip technology and/or sequencing technologies.

Further, step 5 also includes：

Step 5.6, using model prediction evaluation method, calculate the corresponding multiple experimenters' work spies of multiple CDDS values Levy ROC curve, and multiple AUCs corresponding with multiple ROC curve difference；

Step 5.7, the model corresponding to the first ROC curve in multiple ROC curves is defined as into optimal models, institute It is the CDDS a reference values to state the corresponding CDDS values of optimal models, wherein, the Sensitivity and Specificity of first ROC curve is equal More than 85%, and the corresponding AUC of first ROC curve is more than 90%.

It is using the beneficial effect of above-mentioned further scheme：By from calculated many using model prediction evaluation method The model corresponding to the first ROC curve in the corresponding multiple ROC curves of individual CDDS values determines CDDS a reference values, such that it is able to have The accuracy for improving Complex Diseases diagnosis of effect.

Further, the CDDS values are calculated using NB Algorithm.

Further, the CDDS values are calculated using NB Algorithm, including：

Calculate the probability P 1 of the first genetic mutation, the subject when subject suffers from the Complex Diseases and do not suffer from institute The probability P 2 of the first genetic mutation described in when stating Complex Diseases, and the Complex Diseases incidence rate P3 and do not fall ill Probability P 4；

According to described P1, P2, P3 and P4, the CDDS values are calculated.

Further, the CDDS values are calculated according to below equation：

Wherein, N_υThe quantity of the full gene variation included by the modular model.

The advantage of the additional aspect of the present invention will be set forth in part in the description, and partly will become from the following description Obtain substantially, or recognized by present invention practice.

Description of the drawings

In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, below will be to the embodiment of the present invention or prior art The accompanying drawing to be used needed for description is briefly described, it should be apparent that, drawings described below is only the present invention's Some embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, can be with according to this A little accompanying drawings obtain other accompanying drawings.

Fig. 1 is a kind of Complex Diseases prediction and the signal of diagnostic method based on multigroup provided in an embodiment of the present invention Property flow chart；

A kind of Complex Diseases prediction and diagnostic method based on multigroup that Fig. 2 is provided for another embodiment of the present invention Indicative flowchart；

A kind of Complex Diseases prediction and diagnostic method based on multigroup that Fig. 3 is provided for another embodiment of the present invention Indicative flowchart；

Fig. 4 is the first ROC curve schematic diagram provided in an embodiment of the present invention.

Specific embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete Site preparation is described, it is clear that described embodiment is a part of embodiment of the present invention, rather than whole embodiments.Based on this Embodiment in bright, the every other reality that those of ordinary skill in the art are obtained on the premise of creative work is not made Example is applied, should all belong to the scope of protection of the invention.A kind of Complex Diseases prediction and diagnosis based on multigroup as shown in Figure 1 Method 100 includes：

110th, first genomic information of subject is obtained.

Specifically, in this embodiment it is possible to pass through chip technology and/or sequencing technologies the first genomic information of acquisition, But the embodiment of the present invention is not limited thereto.

120th, by the first genomic information and acquired genome mutation site and the relationship information of Complex Diseases Matched, calculate the corresponding Complex Diseases of the first genomic information and differentiate scoring CDDS values.

Specifically, in embodiment, Complex Diseases differentiate scoring (Complex Disease Discrimination Score, CDDS) can be calculated using NB Algorithm, for weighing the probability that subject suffers from institute's detection of complex disease Size.

130th, compare the magnitude relationship of CDDS values and the CDDS a reference values for prestoring, obtain comparative result.

140th, according to comparative result, determine that subject suffers from the probability of institute's detection of complex disease.

A kind of Complex Diseases prediction and diagnostic method based on multigroup provided in above-described embodiment, by will be tested The genomic information of person is matched with acquired genome mutation site with the relationship information of Complex Diseases, and calculating should The corresponding CDDS values of genomic information, and compare CDDS values and determine that subject suffers from institute with the magnitude relationship of CDDS a reference values for prestoring The probability of detection of complex disease, such that it is able to whether subject is carried out prevention and examined in early days with institute's detection of complex disease It is disconnected, the general level of the health of crowd is improved, advance from the process cured the disease to diseases prevention.

Specifically, in this embodiment, in step 140, when comparative result is that CDDS values are more than CDDS a reference values, it is determined that Subject suffers from the probability of institute's detection of complex disease more than 85%.

Or, when comparative result is that CDDS values are less than CDDS a reference values, determine that subject suffers from institute's detection of complex disease Probability be less than 15%.

It should be noted that in this embodiment, the detailed process for calculating CDDS values using NB Algorithm is as follows：

The when calculating subject the probability P 1, subject of the first genetic mutation not suffering from Complex Diseases when suffering from Complex Diseases The probability P 2 of one genetic mutation, and the incidence rate P3 and not incidence rate P4 of Complex Diseases.Further according to below equation meter Calculate CDDS values：

Wherein, N_υThe quantity of the full gene variation included by modular model.First when subject suffers from Complex Diseases The probability of genetic mutationN(A)₁For sum, N (B) under the disease condition of first A form variation₁For first change All types sum under different disease condition.

The probability of the first genetic mutation when subject does not suffer from Complex DiseasesN(A)₁For first A type Sum, N (NB) in the case of the non-diseased of variation₁For all types sum in the case of the non-diseased of first variation.

The incidence rate of Complex DiseasesN_BFor target disease number of the infected, N is total crowd.Complexity disease The not incidence rate P4=1-P3 of disease.

Alternatively, in one embodiment, as shown in Fig. 2 before step 110, method 100 also includes：

150th, according to the healthy population genomic information and target disease crowd's genomic information for obtaining, obtain genome and become Ectopic sites and Complex Diseases relationship information.

Specifically, as shown in figure 3, step 150 can include：

151st, healthy population genomic information and target disease crowd's genomic information are obtained.

Specifically, in this embodiment it is possible to pass through chip technology and/or sequencing technologies acquisition healthy population genome letter Cease and target disease crowd's genomic information, but the embodiment of the present invention is not limited thereto.

152nd, the corresponding multiple models of target disease are built based on multigroup learning.

153rd, healthy population genomic information and target disease crowd genomic information reflected with multiple models respectively Penetrate, obtain healthy population model set and target disease crowd's base model set.

154th, for each model in multiple modules, healthy population model set and target disease crowd's base are calculated respectively Belong to the corresponding CDDS values of the second genomic information of same model in model set, obtain multiple second genomic information difference Corresponding multiple CDDS values.

155th, multiple second genomic informations and multiple CDDS values are stored correspondingly, forms genome mutation site With the relationship information of Complex Diseases, relationship information is the corresponding relation of genomic information and CDDS values.

It should be noted that in this embodiment, the side of the corresponding CDDS values of the second genomic information is calculated in step 154 Method and process with it is identical described in above-mentioned steps 120, it is succinct for description, will not be described here.

Alternatively, in another embodiment, as shown in figure 3, step 150 also includes：

156th, using model prediction evaluation method, the corresponding multiple Receiver Operating Characteristics ROC of multiple CDDS values are calculated bent Line, and multiple AUCs corresponding with multiple ROC curves difference.

157th, the model corresponding to the first ROC curve in multiple ROC curves is defined as into optimal models, optimal models pair The CDDS values answered are CDDS a reference values.

Specifically, in embodiment, as shown in figure 4, the Sensitivity and Specificity of the first ROC curve is all higher than 85%, and The corresponding AUC of first ROC curve is more than 90%.Wherein, ROC curve refers to Receiver operating curve (receiver Operating characteristic curve), reflect the overall target of Sensitivity and Specificity continuous variable, use composition method Disclose Sensitivity and Specificity correlation, it by the way that continuous variable is set out into multiple different critical values, so as to calculate Go out a series of Sensitivity and Specificities, then with sensitiveness as ordinate, specificity be depicted as curve for abscissa.TG-AUC AUC (area under roc curve) is bigger, and diagnostic accuracy is higher.

It should be noted that in this embodiment, corresponding the surveyed Complex Diseases of the first ROC curve are tuberculosis, this In the technical scheme that is intended to be merely illustrative of the present and illustrated example, any limit is not constituted to technical scheme It is fixed.

A kind of Complex Diseases prediction and diagnostic method based on multigroup that above-described embodiment is provided, by from utilizing mould The mould corresponding to the first ROC curve in the corresponding multiple ROC curves of the calculated multiple CDDS values of type prediction and evaluation method Type determines CDDS a reference values, such that it is able to the accuracy for effectively improving Complex Diseases diagnosis.

In addition, the terms "and/or", a kind of only incidence relation of description affiliated partner, expression there may be Three kinds of relations, for example, A and/or B can be represented：Individualism A, while there is A and B, individualism B these three situations.Separately Outward, character "/" herein, typicallys represent forward-backward correlation pair as if a kind of relation of "or".

Those of ordinary skill in the art are it is to be appreciated that the calculation of each example with reference to the embodiments described herein description Method step, can with electronic hardware, computer software or the two be implemented in combination in.

More than, the only specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, any to be familiar with Those skilled in the art the invention discloses technical scope in, various equivalent modifications or replacement can be readily occurred in, These modifications or replacement all should be included within the scope of the present invention.Therefore, protection scope of the present invention should be wanted with right The protection domain asked is defined.

Claims (9)

1. a kind of Complex Diseases prediction and diagnostic method based on multigroup, it is characterised in that comprise the following steps：

Step 1, the first genomic information for obtaining subject；

Step 2, the relevance of first genomic information and acquired genome mutation site and Complex Diseases is believed Breath is matched, and is calculated the corresponding Complex Diseases of first genomic information and is differentiated scoring CDDS values；

Step 3, relatively more described CDDS values and the magnitude relationship of the CDDS a reference values for prestoring, obtain comparative result；

Step 4, according to the comparative result, determine that the subject suffers from the probability of institute's detection of complex disease.

2. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 1, it is characterised in that step 4 include：

When the comparative result is that the CDDS values are more than the CDDS a reference values, determine that the subject suffers from institute's detection of complex The probability of property disease is more than 85%；Or,

When the comparative result is that the CDDS values are less than the CDDS a reference values, determine that the subject suffers from institute's detection of complex Property disease probability be less than 15%.

3. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 1, it is characterised in that in step Before rapid 1, also include：

Step 5, according to obtain healthy population genomic information and target disease crowd's genomic information, obtain genome mutation Site and Complex Diseases relationship information.

4. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 3, it is characterised in that step 5 include：

Step 5.1, the acquisition healthy population genomic information and the target disease crowd genomic information；

Step 5.2, the corresponding multiple models of target disease are built based on multigroup learning；

It is step 5.3, respectively that the healthy population genomic information and the target disease crowd genomic information is more with described Individual model is mapped, and obtains healthy population model set and target disease crowd's base model set；

Step 5.4, for each model in the plurality of module, the healthy population model set and the mesh are calculated respectively Belong to the corresponding CDDS values of the second genomic information of same model in mark sick people's base model set, obtain multiple described the Two gene group information distinguishes corresponding multiple CDDS values；

Step 5.5, multiple second genomic informations and multiple CDDS values are stored correspondingly, form genome The relationship information of variant sites and Complex Diseases, the relationship information is that the genomic information is corresponding with CDDS values Relation.

5. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 4, it is characterised in that described First genomic information, the healthy population genomic information and the target disease crowd genomic information pass through chip skill Art and/or sequencing technologies are obtained.

6. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 4, it is characterised in that step 5 also include：

Step 5.6, using model prediction evaluation method, calculate the corresponding multiple Receiver Operating Characteristics of multiple CDDS values ROC curve, and multiple AUCs corresponding with multiple ROC curve difference；

Step 5.7, the model corresponding to the first ROC curve in multiple ROC curves is defined as optimal models, it is described most The corresponding CDDS values of excellent model are the CDDS a reference values, wherein, the Sensitivity and Specificity of first ROC curve is all higher than 85%, and the corresponding AUC of first ROC curve is more than 90%.

7. the Complex Diseases based on multigroup according to any one of claim 1-6 predict and diagnostic method that it is special Levy and be, using NB Algorithm the CDDS values are calculated.

8. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 7, it is characterised in that utilize NB Algorithm calculates the CDDS values, including：

Calculate the probability P 1 of the first genetic mutation when the subject suffers from the Complex Diseases, the subject do not suffer from it is described multiple The probability P 2 of the first genetic mutation described in during polygamy disease, and the incidence rate P3 and not incidence rate of the Complex Diseases P4；

According to described P1, P2, P3 and P4, the CDDS values are calculated.

9. the Complex Diseases prediction and diagnostic method based on multigroup according to claim 8, it is characterised in that according to Below equation calculates the CDDS values：

$CDDS=\frac{{\Σ}_{i=1}^n{\log }_{10}\left(P1\right)+{\log }_{10}\left(P3\right)-{\Σ}_{i=1}^n{\log }_{10}\left(P2\right)-{\log }_{10}\left(P4\right)}{N_{\mathrm{\υ}}}*1000,$

Wherein, N_υThe quantity of the full gene variation included by the modular model.