CN111312336A - Method and system for establishing biological edge identification system - Google Patents

Abstract

The invention discloses a method and a system for establishing a biological edge identification system, which can simply and efficiently find out the change of the key interaction as the biological identification of the occurrence and development of the disease. The technical scheme is as follows: collecting data with two states; selecting gene pairs whose correlations meet the conditions of significant differences; for gene pairs whose correlations meet the conditions of significant differences, through matrix transformation, the expression value data of the gene pairs are converted into representative correlations. The feature selection algorithm is used to find out the gene pair with the best classification ability in the edge data, and the gene pair with the best classification ability is used as the biological edge identification, so as to establish the biological edge identification system.

Description

Method and system for establishing biological edge identification system

The present invention is a divisional application of the invention patent application with the application number 201410640410.2 and the invention title "Method and System for Establishing Biological Edge Identification System" filed on November 13, 2014.

technical field

The present invention relates to computational systems biology and bioinformatics, in particular to a method and system for processing biological markers.

Background technique

The research of biomarkers has always been an important topic in the field of biomedicine. A successful biomarker can help doctors make an accurate diagnosis or propose an effective treatment plan. Therefore, finding a suitable biomarker is very important to overcome diseases, especially complex diseases. significance.

Human complex disease is a general term for a class of diseases with unclear etiology, many factors, and no effective treatment, such as various types of cancer and diabetes. Since the 1980s, the rapid development of high-throughput biotechnology (such as DNB chips, high-throughput sequencing, etc.) has brought opportunities for the study of complex human diseases.

How to find useful biomarkers from the massive data generated by these technologies is also a major challenge facing the field of biomarker research today. Early research focused on differentially expressed genes or proteins and other biomolecules, and used distinguishing molecules as biomarkers. These methods are simple and intuitive, and have good effects on some simple diseases, but these methods do not consider the difference between molecules. There are complex interactions, and the occurrence of many complex diseases is often caused by changes in the interactions between these molecules, so the application of these methods in complex diseases is not effective.

Because of this, many researchers began to look for biomarkers from the perspective of systems or networks, that is, considering the network composed of various interactions between biomolecules, and using the sub-networks or edge sets with distinguishing ability as biomarkers. There are currently few ideal ways to achieve this.

SUMMARY OF THE INVENTION

A brief summary of one or more aspects is presented below to provide a basic understanding of the aspects. This summary is not an exhaustive overview of all contemplated aspects and is neither intended to identify key or critical elements of all aspects nor attempt to delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.

The purpose of the present invention is to provide a method and system for establishing a biological edge marker system, which can simply and efficiently find changes in key interactions as biological markers for the occurrence and development of diseases.

The technical scheme of the present invention is as follows: the present invention discloses a method for establishing a biological edge identification system, including:

Collect data with two states;

Select the gene pairs whose correlation meets the conditions of significant difference;

For gene pairs whose correlations meet the conditions of significant difference, the expression data of gene pairs are converted into edge data representing correlations through matrix transformation;

The feature selection algorithm is used to find out the gene pair with the best classification ability in the edge data, and the gene pair with the best classification ability is used as the biological edge identification, so as to establish the biological edge identification system.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, the data with dual states includes: normal state data and disease state data, transfer state data and non-transfer state data, data with drug resistance state and data without Data on drug resistance status.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, the data type of the data with two states includes expression profile or abundance profile data of gene pairs.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, after the step of collecting data with two states, the method further includes:

The data is preprocessed to remove genes whose expression mean is lower than the set value or whose coefficient of variation is higher than the set value.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, in the step of selecting gene pairs whose correlations meet the conditions of significant difference, the correlation coefficients of the gene pairs in two states are calculated, according to the two states The absolute value of the difference of the correlation coefficient is compared with the threshold value to determine whether the correlation meets the conditions of significant difference.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, for the gene pairs whose correlations meet the conditions of significant difference, the expression value data of the gene pairs are converted into edge data representing the correlation through matrix transformation. In step, the expression value data of gene pairs are in matrix form:

Wherein, x _ij represents the value of the expression profile or the value of the abundance profile of the j-th sample of biomolecule i in the first state of the two states, and y _ij represents the second state of the biomolecule i in the two states The numerical value of the expression profile or the numerical value of the abundance profile of the jth sample in the state;

The process of matrix transformation is:

For a given gene pair u and v, do the following transformation:

分别是基因对u和v在第一状态和第二状态下的表达谱的数值或丰度谱的数值的均值，Sx_u，Sx_v，Sy_u，Sy_v分别是基因对u和v在第一状态下和第二状态下的方差，k₁，k₂为校正系数，所有相关性符合显著差异条件的基因对得到的<u,v>_N和<u,v>_D所组成的矩阵就是基因对对应的边数据，边数据代表该基因对在不同状态下的相关性，每一个基因对由边数据里的两个对偶的变量或特征所刻画。Among them, <u, v> _N and <u, v> _D refer to the edge features of the gene pair u, v in the first state and the second state, respectively,

are the mean values of expression profiles or abundance profiles of gene pairs u and v in the first and second states, respectively, Sx _u , Sx _v , Sy _u , Sy _v are the gene pairs u and v in the first and second states, respectively. The variance in the first state and the second state, k ₁ , k ₂ are correction coefficients, and the matrix composed of <u,v> _N and <u,v> _D obtained from all gene pairs whose correlations meet the conditions of significant difference is The edge data corresponding to the gene pair, the edge data represents the correlation of the gene pair in different states, and each gene pair is characterized by two dual variables or features in the edge data.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, the values of the correction coefficients k ₁ and k ₂ are both 1.

According to an embodiment of the method for establishing a biological edge identification system of the present invention, the feature selection algorithm includes Sequential Forward Floating Selection (SFFS) and Support Vector Machine (SVM) in machine learning.

The present invention discloses a biological edge identification system, comprising:

An information collection module that collects data with two states;

The gene pair selection module selects gene pairs whose correlations meet the conditions of significant difference;

The edge data acquisition module converts the expression value data of the gene pair into edge data representing the correlation through matrix transformation for gene pairs whose correlations meet the conditions of significant difference;

The biological edge identification building module uses the feature selection algorithm to find out the gene pair with the best classification ability in the edge data, and uses the gene pair with the best classification ability as the biological edge identification, thereby establishing the biological edge identification system.

According to an embodiment of the biological edge identification system of the present invention, the data with dual states includes: normal state data and disease state data, transfer state data and non-transfer state data, drug resistance state data and drug resistance state data The data.

According to an embodiment of the biological edge identification system of the present invention, the data type of the data with two states includes expression profile or abundance profile data of gene pairs.

According to an embodiment of the biometric edge identification system of the present invention, after the information collection module is further connected:

The preprocessing module preprocesses the data to remove genes whose expression mean is lower than the set value or whose coefficient of variation is higher than the set value.

According to an embodiment of the biological edge identification system of the present invention, in the gene pair selection module, the correlation coefficient of the gene pair under two states is calculated, and the correlation is determined according to the comparison between the absolute value of the difference of the correlation coefficient under the two states and the threshold value. Whether the sex meets the conditions of significant difference.

According to an embodiment of the biological edge identification system of the present invention, in the edge data acquisition module, the expression value data of the gene pair is in the form of a matrix:

Wherein, x _ij represents the value of the expression profile or the value of the abundance profile of the j-th sample of biomolecule i in the first state of the two states, and y _ij represents the second state of the biomolecule i in the two states The numerical value of the expression profile or the numerical value of the abundance profile of the jth sample in the state;

The process of matrix transformation is:

For a given gene pair u and v, do the following transformation:

分别是基因对u和v在第一状态和第二状态下的表达谱的数值或丰度谱的数值的均值，Sx_u，Sx_v，Sy_u，Sy_v分别是基因对u和v在第一状态下和第二状态下的方差，k₁，k₂为校正系数，所有相关性符合显著差异条件的基因对得到的<u,v>_N和<u,v>_D所组成的矩阵就是基因对对应的边数据，边数据代表该基因对在不同状态下的相关性，每一个基因对由边数据里的两个对偶的变量或特征所刻画。Among them, <u, v> _N and <u, v> _D refer to the edge features of the gene pair u, v in the first state and the second state, respectively,

are the mean values of expression profiles or abundance profiles of gene pairs u and v in the first and second states, respectively, Sx _u , Sx _v , Sy _u , Sy _v are the gene pairs u and v in the first and second states, respectively. The variance in the first state and the second state, k ₁ , k ₂ are correction coefficients, and the matrix composed of <u,v> _N and <u,v> _D obtained from all gene pairs whose correlations meet the conditions of significant difference is The edge data corresponding to the gene pair, the edge data represents the correlation of the gene pair in different states, and each gene pair is characterized by two dual variables or features in the edge data.

According to an embodiment of the biological edge identification system of the present invention, the values of the correction coefficients k ₁ and k ₂ are both 1.

According to an embodiment of the biological edge identification system of the present invention, the feature selection algorithm in the biological edge identification building module includes a cyclic addition and subtraction method (Sequential Forward Floating Selection, SFFS) and a Support Vector Machine (SVM) in machine learning. ).

Compared with the prior art, the present invention has the following beneficial effects: traditional biological markers use the expression of one or more genes (also called molecules) to distinguish different states, while biological edge markers use the correlation between pairs of genes to distinguish different states. Because there are intricate interaction networks between molecules in living organisms, changes in these interactions are often the key factors leading to the development of complex diseases. Therefore, biological edge markers have stronger biological significance than traditional biological markers. Identifying these key interactions as biomarkers for the occurrence and development of diseases can better reveal the underlying mechanisms.

Description of drawings

FIG. 1 shows a flow chart of a first embodiment of a method for establishing a biological edge identification system of the present invention.

FIG. 2 shows a flow chart of the second embodiment of the method for establishing the biological edge identification system of the present invention.

Figure 3 shows a schematic diagram of a first embodiment of the biometric edge identification system of the present invention.

Figure 4 shows a schematic diagram of a second embodiment of the biometric edge identification system of the present invention.

FIG. 5 shows a schematic flowchart of the establishment and application of the biological edge identification system.

Detailed ways

The above-described features and advantages of the present invention can be better understood after reading the detailed description of the embodiments of the present disclosure in conjunction with the following drawings. In the drawings, components are not necessarily drawn to scale and components with similar related characteristics or features may have the same or similar reference numbers.

FIG. 1 shows the flow of the first embodiment of the method for establishing the biological edge identification system of the present invention. Referring to FIG. 1 , the implementation steps of the method for establishing a biological edge identification system in this embodiment are as follows.

Step S11: Collect data with two states.

Data with dual states includes: normal state data and disease state data, metastatic state data and non-metastatic state data, data with drug resistance state and data without drug resistance state. The data types of data with two states include gene pair expression profile or abundance profile data, including microarray, protein profile, metabolite mass spectrometry, epigenetic profile data, etc.

Step S12: Select gene pairs whose correlations meet the conditions of significant difference.

In this step, the correlation coefficient of the gene pair under the two states is calculated, and whether the correlation meets the condition of significant difference is determined according to the comparison of the absolute value of the difference of the correlation coefficient under the two states and the threshold value.

For example, if the correlation coefficient between gene 1 and gene 2 is 0.8 under normal conditions and -0.6 in disease state, then the absolute value of their correlation difference is 1.4. Assuming a threshold of 0.8, gene 1 and gene 2 can be determined. Gene pairs with significantly different correlations.

Step S13: For gene pairs whose correlations meet the conditions of significant difference, the expression value data of the gene pairs are converted into edge data representing correlations through matrix transformation.

This step is the key to the implementation of the present invention. The expression value data for gene pairs are in matrix form:

Wherein, x _ij represents the value of the expression profile or the value of the abundance profile of the j-th sample of biomolecule i in the first state of the two states, and y _ij represents the second state of the biomolecule i in the two states The numerical value of the expression profile or the numerical value of the abundance profile of the jth sample in the state.

The process of matrix transformation is:

For a given gene pair u and v, do the following transformation:

分别是基因对u和v在第一状态和第二状态下的表达谱的数值或丰度谱的数值的均值，Sx_u，Sx_v，Sy_u，Sy_v分别是基因对u和v在第一状态下和第二状态下的方差，k₁，k₂为校正系数(其中校正系数k₁，k₂的取值均为1)，所有相关性符合显著差异条件的基因对得到的<u,v>_N和<u,v>_D所组成的矩阵就是基因对对应的边数据，边数据代表该基因对在不同状态下的相关性，每一个基因对由边数据里的两个对偶的变量或特征所刻画。每个基因对可以根据上述矩阵转换算出一个2*(m+n)的小矩阵，把这些小矩阵按行堆叠在一起得到的大矩阵就是所谓的边数据。这个边数据中每一行代表一个基因对，每一列代表一个样本。Among them, <u, v> _N and <u, v> _D refer to the edge features of the gene pair u, v in the first state and the second state, respectively,

are the mean values of expression profiles or abundance profiles of gene pairs u and v in the first and second states, respectively, Sx _u , Sx _v , Sy _u , Sy _v are the gene pairs u and v in the first and second states, respectively. The variance in the first state and the second state, k ₁ , k ₂ are the correction coefficients (where the correction coefficients k ₁ , k ₂ are both 1), and all the gene pairs whose correlations meet the conditions of significant difference get <u ,v> _N and <u,v> _D matrix is the edge data corresponding to the gene pair, the edge data represents the correlation of the gene pair in different states, each gene pair is composed of two duals in the edge data. variable or characteristic. For each gene pair, a small matrix of 2*(m+n) can be calculated according to the above matrix transformation, and the large matrix obtained by stacking these small matrices together in rows is the so-called edge data. Each row in this edge data represents a gene pair, and each column represents a sample.

Step S14: Find out the gene pair with the best classification ability in the edge data by applying the feature selection algorithm, and use the gene pair with the best classification ability as the biological edge identification, thereby establishing the biological edge identification system.

In this step, the feature selection algorithm includes Sequential Forward Floating Selection (SFFS) and Support Vector Machine (SVM) in machine learning.

FIG. 2 shows the flow of the second embodiment of the method for establishing the biometric edge identification system of the present invention. Referring to FIG. 2 , the implementation steps of the method for establishing a biological edge identification system in this embodiment are as follows.

Step S21: Collect data with two states.

Data with dual states includes: normal state data and disease state data, metastatic state data and non-metastatic state data, data with drug resistance state and data without drug resistance state. The data types of data with two states include gene pair expression profile or abundance profile data, including microarray, protein profile, metabolite mass spectrometry, epigenetic profile data, etc.

Step S22: Preprocess the data to remove genes whose expression mean is lower than the set value or whose coefficient of variation is higher than the set value.

Step S23: Select gene pairs whose correlations meet the conditions of significant difference.

In this step, the correlation coefficient of the gene pair under the two states is calculated, and whether the correlation meets the condition of significant difference is determined according to the comparison of the absolute value of the difference of the correlation coefficient under the two states and the threshold value.

For example, if the correlation coefficient between gene 1 and gene 2 is 0.8 under normal conditions and -0.6 in disease state, then the absolute value of their correlation difference is 1.4. Assuming a threshold of 0.8, gene 1 and gene 2 can be determined. Gene pairs with significantly different correlations.

Step S24: For gene pairs whose correlations meet the conditions of significant difference, transform the expression value data of the gene pairs into edge data representing the correlations through matrix transformation.

This step is the key to the implementation of the present invention. The expression value data for gene pairs are in matrix form:

Wherein, x _ij represents the value of the expression profile or the value of the abundance profile of the j-th sample of biomolecule i in the first state of the two states, and y _ij represents the second state of the biomolecule i in the two states The numerical value of the expression profile or the numerical value of the abundance profile of the jth sample in the state.

The process of matrix transformation is:

For a given gene pair u and v, do the following transformation:

分别是基因对u和v在第一状态和第二状态下的表达谱的数值或丰度谱的数值的均值，Sx_u，Sx_v，Sy_u，Sy_v分别是基因对u和v在第一状态下和第二状态下的方差，k₁，k₂为校正系数(其中校正系数k₁，k₂的取值均为1)，所有相关性符合显著差异条件的基因对得到的<u,v>_N和<u,v>_D所组成的矩阵就是基因对对应的边数据，边数据代表该基因对在不同状态下的相关性，每一个基因对由边数据里的两个对偶的变量或特征所刻画。每个基因对可以根据上述矩阵转换算出一个2*(m+n)的小矩阵，把这些小矩阵按行堆叠在一起得到的大矩阵就是所谓的边数据。这个边数据中每一行代表一个基因对，每一列代表一个样本。Among them, <u, v> _N and <u, v> _D refer to the edge features of the gene pair u, v in the first state and the second state, respectively,

are the mean values of expression profiles or abundance profiles of gene pairs u and v in the first and second states, respectively, Sx _u , Sx _v , Sy _u , Sy _v are the gene pairs u and v in the first and second states, respectively. The variance in the first state and the second state, k ₁ , k ₂ are the correction coefficients (where the correction coefficients k ₁ , k ₂ are both 1), and all the gene pairs whose correlations meet the conditions of significant difference get <u ,v> _N and <u,v> _D matrix is the edge data corresponding to the gene pair, the edge data represents the correlation of the gene pair in different states, each gene pair is composed of two duals in the edge data. variable or characteristic. For each gene pair, a small matrix of 2*(m+n) can be calculated according to the above matrix transformation, and the large matrix obtained by stacking these small matrices together in rows is the so-called edge data. Each row in this edge data represents a gene pair, and each column represents a sample.

Step S25 : find out the gene pair with the best classification ability in the edge data by applying the feature selection algorithm, and use the gene pair with the best classification ability as the biological edge identification, thereby establishing the biological edge identification system.

In this step, the feature selection algorithm includes Sequential Forward Floating Selection (SFFS) and Support Vector Machine (SVM) in machine learning.

Figure 3 shows the principle of the first embodiment of the biometric edge identification system of the present invention. Referring to FIG. 3 , the biological edge identification system in this embodiment includes: an information collection module 11 , a gene pair selection module 12 , an edge data acquisition module 13 , and a biological edge identification establishment module 14 .

The connection between these modules is that the information collection module 11 is connected to the gene pair selection module 12 , the gene pair selection module 12 is connected to the edge data acquisition module 13 , and the edge data acquisition module 13 is connected to the biological edge identification establishment module 14 .

The information collection module 11 collects data with two states. Data with dual states includes: normal state data and disease state data, metastatic state data and non-metastatic state data, data with drug resistance state and data without drug resistance state. The data types of data with two states include gene pair expression profile or abundance profile data, including microarray, protein profile, metabolite mass spectrometry, epigenetic profile data, etc.

The gene pair selection module 12 selects gene pairs whose correlations meet the conditions of significant difference. The gene pair selection module 12 calculates the correlation coefficient of the gene pair under two states, and determines whether the correlation meets the condition of significant difference according to the comparison between the absolute value of the difference of the correlation coefficient under the two states and the threshold value.

For example, if the correlation coefficient between gene 1 and gene 2 is 0.8 under normal conditions and -0.6 in disease state, then the absolute value of their correlation difference is 1.4. Assuming a threshold of 0.8, gene 1 and gene 2 can be determined. Gene pairs with significantly different correlations.

The edge data acquisition module 13 converts the expression value data of the gene pair into edge data representing the correlation through matrix transformation for the gene pairs whose correlations meet the conditions of significant difference.

The edge data acquisition module 13 is the key to implementing the present invention. The expression value data for gene pairs are in matrix form:

Wherein, x _ij represents the value of the expression profile or the value of the abundance profile of the j-th sample of biomolecule i in the first state of the two states, and y _ij represents the second state of the biomolecule i in the two states The numerical value of the expression profile or the numerical value of the abundance profile of the jth sample in the state.

The process of matrix transformation is:

For a given gene pair u and v, do the following transformation:

分别是基因对u和v在第一状态和第二状态下的表达谱的数值或丰度谱的数值的均值，Sx_u，Sx_v，Sy_u，Sy_v分别是基因对u和v在第一状态下和第二状态下的方差，k₁，k₂为校正系数(其中校正系数k₁，k₂的取值均为1)，所有相关性符合显著差异条件的基因对得到的<u,v>_N和<u,v>_D所组成的矩阵就是基因对对应的边数据，边数据代表该基因对在不同状态下的相关性，每一个基因对由边数据里的两个对偶的变量或特征所刻画。每个基因对可以根据上述矩阵转换算出一个2*(m+n)的小矩阵，把这些小矩阵按行堆叠在一起得到的大矩阵就是所谓的边数据。这个边数据中每一行代表一个基因对，每一列代表一个样本。Among them, <u, v> _N and <u, v> _D refer to the edge features of the gene pair u, v in the first state and the second state, respectively,

are the mean values of expression profiles or abundance profiles of gene pairs u and v in the first and second states, respectively, Sx _u , Sx _v , Sy _u , Sy _v are the gene pairs u and v in the first and second states, respectively. The variance in the first state and the second state, k ₁ , k ₂ are the correction coefficients (where the correction coefficients k ₁ , k ₂ are both 1), and all the gene pairs whose correlations meet the conditions of significant difference get <u ,v> _N and <u,v> _D matrix is the edge data corresponding to the gene pair, the edge data represents the correlation of the gene pair in different states, each gene pair is composed of two duals in the edge data. variable or characteristic. For each gene pair, a small matrix of 2*(m+n) can be calculated according to the above matrix transformation, and the large matrix obtained by stacking these small matrices together in rows is the so-called edge data. Each row in this edge data represents a gene pair, and each column represents a sample.

The biological edge identification building module 14 applies the feature selection algorithm to find out the gene pair with the best classification ability in the edge data, and uses the gene pair with the best classification ability as the biological edge identification, thereby establishing the biological edge identification system. Feature selection algorithms include Sequential Forward Floating Selection (SFFS) and Support Vector Machine (SVM) in machine learning.

Figure 4 shows the principle of the second embodiment of the biometric edge identification system of the present invention. Referring to FIG. 4 , the biological edge identification system in this embodiment includes: an information collection module 21 , a preprocessing module 22 , a gene pair selection module 23 , an edge data acquisition module 24 , and a biological edge identification establishment module 25 .

The connection relationship between these modules is that the information collection module 21 is connected to the preprocessing module 22, the preprocessing module 22 is connected to the gene pair selection module 23, the gene pair selection module 23 is connected to the side data acquisition module 24, and the side data acquisition module 24 Afterwards, the biological edge identification establishing module 25 is connected.

The information collection module 21 collects data with two states. Data with dual states includes: normal state data and disease state data, metastatic state data and non-metastatic state data, data with drug resistance state and data without drug resistance state. The data types of data with two states include gene pair expression profile or abundance profile data, including microarray, protein profile, metabolite mass spectrometry, epigenetic profile data, etc.

The preprocessing module 22 preprocesses the data, and removes genes whose expression mean value is lower than the set value or the variation coefficient is higher than the set value, so as to reduce the influence of noise on the result.

The gene pair selection module 23 selects gene pairs whose correlations meet the conditions of significant difference. The gene pair selection module 23 calculates the correlation coefficient of the gene pair under two states, and determines whether the correlation meets the significant difference condition according to the comparison between the absolute value of the difference of the correlation coefficient under the two states and the threshold value.

For example, if the correlation coefficient between gene 1 and gene 2 is 0.8 under normal conditions and -0.6 in disease state, then the absolute value of their correlation difference is 1.4. Assuming a threshold of 0.8, gene 1 and gene 2 can be determined. Gene pairs with significantly different correlations.

The edge data acquisition module 24 converts the expression value data of the gene pairs into edge data representing the correlation through matrix transformation for gene pairs whose correlations meet the conditions of significant difference.

The edge data acquisition module 24 is the key to implementing the present invention. The expression value data for gene pairs are in matrix form:

Wherein, x _ij represents the value of the expression profile or the value of the abundance profile of the j-th sample of biomolecule i in the first state of the two states, and y _ij represents the second state of the biomolecule i in the two states The numerical value of the expression profile or the numerical value of the abundance profile of the jth sample in the state.

The process of matrix transformation is:

For a given gene pair u and v, do the following transformation:

分别是基因对u和v在第一状态和第二状态下的表达谱的数值或丰度谱的数值的均值，Sx_u，Sx_v，Sy_u，Sy_v分别是基因对u和v在第一状态下和第二状态下的方差，k₁，k₂为校正系数(其中校正系数k₁，k₂的取值均为1)，所有相关性符合显著差异条件的基因对得到的<u,v>_N和<u,v>_D所组成的矩阵就是基因对对应的边数据，边数据代表该基因对在不同状态下的相关性，每一个基因对由边数据里的两个对偶的变量或特征所刻画。每个基因对可以根据上述矩阵转换算出一个2*(m+n)的小矩阵，把这些小矩阵按行堆叠在一起得到的大矩阵就是所谓的边数据。这个边数据中每一行代表一个基因对，每一列代表一个样本。Among them, <u, v> _N and <u, v> _D refer to the edge features of the gene pair u, v in the first state and the second state, respectively,

are the mean values of expression profiles or abundance profiles of gene pairs u and v in the first and second states, respectively, Sx _u , Sx _v , Sy _u , Sy _v are the gene pairs u and v in the first and second states, respectively. The variance in the first state and the second state, k ₁ , k ₂ are the correction coefficients (where the correction coefficients k ₁ , k ₂ are both 1), and all the gene pairs whose correlations meet the conditions of significant difference get <u ,v> _N and <u,v> _D matrix is the edge data corresponding to the gene pair, the edge data represents the correlation of the gene pair in different states, each gene pair is composed of two duals in the edge data. variable or characteristic. For each gene pair, a small matrix of 2*(m+n) can be calculated according to the above matrix transformation, and the large matrix obtained by stacking these small matrices together in rows is the so-called edge data. Each row in this edge data represents a gene pair, and each column represents a sample.

The biological edge identification building module 25 applies the feature selection algorithm to find out the gene pair with the best classification ability in the edge data, and uses the gene pair with the best classification ability as the biological edge identification, thereby establishing a biological edge identification system. Feature selection algorithms include Sequential Forward Floating Selection (SFFS) and Support Vector Machine (SVM) in machine learning.

FIG. 5 also shows a schematic flow of the establishment and application of an example of the biometric edge identification system. With the biological edge identification system established by the present invention, a corresponding classifier or diagnostic model can be established. Based on this diagnostic model, for the sample to be tested, the boundary value corresponding to the biological edge identification is calculated according to the matrix transformation as the diagnostic model. Input data, and then judge the status of the sample to be tested according to the output results of the model, that is, whether it is sick or whether cancer metastasis occurs.

Taking gene expression data as an example, existing methods mainly select genes with the greatest discriminating ability from differentially expressed genes as biomarkers, and it is unknown whether these biomarkers have interactions. The present invention focuses on genes with differences in interaction, and finds the gene pair with the greatest discriminative ability as a biological marker, which is called a biological edge marker. causes of disease development.

Although the above-described methods are illustrated and described as a series of acts for simplicity of explanation, it should be understood and appreciated that these methods are not limited by the order of the acts, as some acts may occur in a different order in accordance with one or more embodiments and/or occur concurrently with other actions from or not shown and described herein but understood by those skilled in the art.

Those skilled in the art will further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

The various illustrative logic blocks, modules, and circuits described in connection with the embodiments disclosed herein may be implemented using general purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other Programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein are implemented or performed. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in the ASIC. The ASIC may reside in the user terminal. In the alternative, the processor and storage medium may reside in the user terminal as discrete components.

In one or more exemplary embodiments, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage medium can be any available medium that can be accessed by a computer. By way of example and not limitation, such computer-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage, or other magnetic storage devices, or can be used to carry or store instructions or data structures in the form of Any other medium that conforms to program code and that can be accessed by a computer. Any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a web site, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave , then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc as used herein includes compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and blu-ray disc, where disks are often reproduced magnetically data, and discs reproduce the data optically with a laser. Combinations of the above should also be included within the scope of computer-readable media.

The previous description of the present disclosure is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to the present disclosure will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of the present disclosure. Thus, the present disclosure is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (10)

1. A method for establishing a biological edge identification system comprises the following steps:

collecting data having a dual state;

selecting a gene pair with correlation meeting a significant difference condition;

for the gene pairs whose correlation meets the condition of significant difference, the expression value data of the gene pairs is converted into edge data representing the correlation by matrix transformation, wherein the expression value data of the gene pairs is in the form of a matrix:

wherein x is_ijA value representing the expression profile or abundance profile of the j-th sample of the biomolecule i in the first of the two states, y_ijA value representing the expression profile or abundance profile of the sample j in the second of the two states of the biomolecule i;

the process of matrix transformation is:

for a given gene pair u and v, the following transformations are made:

wherein,<u,v>_Nand<u,v>_Dthe edge characteristics of the gene pair u, v in the first state and the second state respectively,

the mean of the values of the expression or abundance profiles of the gene pairs u and v in the first and second states, Sx_u，Sx_v，Sy_u，Sy_vThe variances, k, of the gene pairs u and v in the first and second states, respectively₁，k₂All the genes with correlation meeting the condition of significant difference are obtained for correcting the coefficient and taking the value of 1<u,v>_NAnd<u,v>_Dthe formed matrix is the edge data corresponding to the gene pair, the edge data represents the relativity of the gene pair under different states, and each gene pair is described by two dual variables or characteristics in the edge data; and

and (3) finding out the gene pair with the best classification capability in the side data by using an applied feature selection algorithm comprising a cyclic addition and subtraction method and a support vector machine in machine learning, and taking the gene pair with the best classification capability as the biological side identifier, thereby establishing a biological side identifier system.

2. The method for establishing a biometric edge identification system as claimed in claim 1, wherein the data having two states comprises: normal state data and disease state data, metastatic state data and non-metastatic state data, data for a drug resistant state and data for a drug non-resistant state.

3. The method for establishing a biological edge marking system as claimed in claim 1, wherein the data type of the data with two states comprises expression profile or abundance profile data of gene pairs.

4. The method for establishing a biometric edge identification system as claimed in claim 1, further comprising, after the step of collecting data having two states:

and (3) preprocessing the data, and removing genes of which the expression mean value is lower than a set value or the variation coefficient is higher than a set value.

5. The method for establishing a biometric edge identification system according to claim 1, wherein in the step of selecting the pair of genes whose correlations satisfy the condition of significant difference, the correlation coefficient of the pair of genes in the two states is calculated, and whether the correlations satisfy the condition of significant difference is determined according to a comparison between an absolute value of the difference of the correlation coefficients in the two states and a threshold.

6. A biometric edge identification system comprising:

an information collection module that collects data having a dual state;

the gene pair selection module selects a gene pair with correlation meeting the significant difference condition;

and the side data acquisition module is used for converting the expression value data of the gene pair into side data representing the correlation through matrix transformation for the gene pair of which the correlation meets the significant difference condition, wherein the expression value data of the gene pair is in a matrix form:

wherein x is_ijA value representing the expression profile or abundance profile of the j-th sample of the biomolecule i in the first of the two states, y_ijA value representing the expression profile or abundance profile of the sample j in the second of the two states of the biomolecule i;

the process of matrix transformation is:

for a given gene pair u and v, the following transformations are made:

wherein,<u,v>_Nand<u,v>_Dthe edge characteristics of the gene pair u, v in the first state and the second state respectively,

the mean of the values of the expression or abundance profiles of the gene pairs u and v in the first and second states, Sx_u，Sx_v，Sy_u，Sy_vThe variances, k, of the gene pairs u and v in the first and second states, respectively₁，k₂All the genes with correlation meeting the condition of significant difference are obtained for correcting the coefficient and taking the value of 1<u,v>_NAnd<u,v>_Dthe formed matrix is the edge data and the edge number corresponding to the gene pairsEach gene pair is characterized by two dual variables or characteristics in the edge data according to the relevance of the gene pair under different states;

and the biological edge identification establishing module is used for finding out the gene pair with the best classification capability in the edge data by using an applied characteristic selection algorithm comprising a cyclic addition and subtraction method and a support vector machine in machine learning, and taking the gene pair with the best classification capability as the biological edge identification, thereby establishing a biological edge identification system.

7. The biometric edge identification system of claim 6, wherein the data having a two-state comprises: normal state data and disease state data, metastatic state data and non-metastatic state data, data for a drug resistant state and data for a drug non-resistant state.

8. The biometric edge identification system of claim 6, wherein the data type of the data having two states comprises expression profile or abundance profile data of a gene pair.

9. The biometric edge identification system according to claim 6, further comprising, after the information collection module:

and the preprocessing module is used for preprocessing the data and removing genes of which the expression mean value is lower than a set value or the variation coefficient is higher than the set value.

10. The biometric edge identification system of claim 6, wherein in the pair selection module, the correlation coefficients of the pair of the biometric edges in the two states are calculated, and whether the correlation meets the condition of significant difference is determined according to the comparison between the absolute value of the difference of the correlation coefficients in the two states and a threshold.

Images