KR102187594B1 - Multi-omics data processing apparatus and method for discovering new drug candidates - Google Patents

Abstract

A data processing method for discovering a new drug candidate substance in a data processing apparatus according to an embodiment of the present invention includes: receiving at least some of a plurality of ohmic levels forming an ohmic through a user interface; Receiving at least some types of correlations among the plurality of types of correlations forming the ohmic through a user interface; Selecting a DB for the at least some of the ohmic levels and a DB for the at least some types of correlations from an ohmics DB including data for each ohmic level and data for each type of correlation; Generating a first matrix consisting of a DB related to the at least part of the ohmic level and a DB related to the type of the at least part of the correlation; Receiving a predetermined search word through a user interface; Extracting a plurality of biological entities related to the predetermined search word and a correlation degree between the plurality of biological entities from the at least part of the ohmic level DB and the at least part of the DB related to the type of correlation; And generating a multiomics network in which a plurality of nodes including the plurality of biological entities are connected according to a degree of correlation between the plurality of biological entities, and some of the plurality of biological entities Included in different ohmic levels.

Description

Multi-omics data processing device and method for discovering new drug candidates {MULTI-OMICS DATA PROCESSING APPARATUS AND METHOD FOR DISCOVERING NEW DRUG CANDIDATES}

The present invention relates to a multiomics data processing apparatus and method for discovering new drug candidates, and more particularly, to a data processing device for generating a multiomics network having a hierarchical structure from an ohmics DB to discover new drug candidates, and It's about how.

It is known that it takes a total of 15 years on average to develop a new drug, and costs 2 to 3 trillion won. Among them, it is known that it takes about 6 years to discover new drug candidates before the preclinical trial.

In general, in order to discover new drug candidates, which is the first step in the pipeline to develop new drugs, a large number of specialized research personnel search for enormous amounts of information one by one, and from this, infer associations between major biological entities. Going through the process.

According to the Life Intelligence Consortium (2017), which was recently launched in Japan, when using artificial intelligence technology to develop a new drug, the time required to develop a new drug can be reduced to about 40%, and the cost is about It is predicted that it can be reduced to 50% level.

On the other hand, omics, also called somatics, is a term that refers to the entire collection of biomolecules, cells, tissues, organs, etc., including genomes, such as genomics and proteomics. , Metabolomics, etc. 1 shows the hierarchical structure of the body. Recently, the concept of multiomics, which means a comprehensive and integrated analysis between different ohmic levels, has been introduced, and it is necessary to utilize such a multiomics network to develop a new drug with a high hit rate.

However, a specific method for efficiently generating a multi-omics network using ohmic data has not been developed.

The technical problem to be solved by the present invention is to provide a data processing apparatus and method for discovering new drug candidates.

Another technical problem to be solved by the present invention relates to an apparatus and method for generating a multi-omics network from an omics database (DB).

A data processing method for discovering a new drug candidate substance performed in a data processing apparatus includes receiving at least some of the ohmic levels from among a plurality of ohmic levels constituting an ohmic through a user interface, and receiving a plurality of ohmic levels constituting the ohmic. Receiving at least some of the types of correlations through a user interface, and from the ohmics DB including data for each ohmic level and data for each type of correlation, a DB on the at least some of the ohmics levels, and Selecting a DB for the at least some types of correlations, generating a first matrix consisting of a DB for at least some of the ohmic levels and a DB for the at least some types of correlations, a user interface Receiving a search word input through the first matrix, extracting biological entities belonging to an ohmic level different from the ohmic level of the search word and related to the search word from the first matrix, and extracting a correlation between the search word and the biological entities Generating a multiomics network in which nodes representing the search word and the biological entities are connected according to a degree of correlation between the search word and the biological entities or between the biological entities, the multiomics Generating a graph theory index for each of the nodes of the network, and having a connection relationship between different ohmic levels in the multiomics network using some nodes extracted using the graph theory index among the nodes Including the step of extracting some pathways, some of the biological entities are included in different ohmic levels from the other biological entities, and the search word is a gene name, protein name, metabolite name, symptom name, disease name, compound Includes at least one of name and drug name, and the biological entity includes at least one of genes, proteins, metabolites, symptoms, diseases, compounds, and drugs, and the category of correlation is participate ), covariate, regulate, associate, bind, upregulate, resemble, treat, downregulates, palliate , Include, and expression, and in the first matrix, the at least some ohmic levels are arranged on each of a horizontal axis and a vertical axis, and the type of correlation is at a point where the horizontal axis and the vertical axis intersect. Is generated to be displayed, and the graph theory index includes a shortest path between nodes for at least one of the nodes constituting the multiomics network, a clustering coefficient for each node, a centrality coefficient for each node, and a connection line between the nodes The weight of the connection line is set differently according to the category of the degree of correlation indicated, and the shortest path between nodes is calculated by reflecting the set weight, and the nodes constituting the partial path are nodes of the multiomics network. The standard score for at least one of the shortest path between nodes for each, the clustering coefficient for each node, and the centrality coefficient for each node is generated by deleting a connection line between a node having a value less than a threshold value and a node having a value less than the threshold value, and the standard A score is a value obtained by dividing a difference between an index value of the graph theory index for each node of the multiomics network and an average index value of the graph theory index for nodes of the multiomics network by a standard error.

The step of extracting the partial paths includes randomly mixing all the connecting lines constituting the multiomics network and then calculating the standard score for each of the nodes of the multiomics network, and the number of randomly mixing is More than 1000 times..

The data processing method may further include displaying some paths having a connection relationship between different ohmic levels in the multiomics network in a hierarchical structure.

The category of the correlation diagram may further include at least one of interaction, cause, present, and localize.

The plurality of ohmic levels may include at least some of a gene level, a protein level, a metabolite level, a symptom level, a disease level, a compound level, a drug level, and a side effect level.

The generating of the multiomics network includes generating a second matrix consisting of the biological entities and the correlation between the biological entities, and connecting the biological entities with the correlation between the biological entities. In the second matrix, the biological entities are sequentially arranged on a horizontal axis and a vertical axis according to a hierarchical structure of an ohmic level, and a correlation between the biological entities may be displayed at a point where the horizontal axis and the vertical axis intersect. have.

The data processing apparatus for discovering new drug candidates receives at least some of the ohmic levels of a plurality of ohmics levels constituting the ohmics, and determines at least some of the types of correlations of the plurality of types of the ohmics. Selecting the at least some of the Omix level DB and the at least some of the Omics level DB and the at least some of the mutual relationship type DB from the Omix DB including the input user interface unit, Omix level data and correlation type data, , A DB extraction unit for generating a first matrix consisting of a DB for the at least part of the ohmic level and a DB for the type of the at least part of the correlation, and an error of the search word input from the first matrix through the user interface unit. Extract biological entities belonging to an ohmic level different from the mix level and related to the search word, extract a correlation between the search word and the biological entities, and select a plurality of nodes representing the search word and the biological entities with the search word A data generation unit for generating a multiomics network connected according to the degree of correlation between the biological entities or the degree of correlation between the biological entities, and generating a graph theory index for each of the nodes of the multiomics network A data processing unit and a data refiner for extracting some paths having a connection relationship between different ohmic levels in the multiomics network by using some nodes extracted using the graph theory index among the nodes, the Some of the biological entities are included in an ohmic level different from the other biological entities, and the search word includes at least one of a gene name, a protein name, a metabolite name, a symptom name, a disease name, a compound name, and a drug name, and the The biological entity includes at least one of genes, proteins, metabolites, symptoms, diseases, compounds, and drugs, and the category of correlation is participate, covariate, regulate, and linkage. Includes associate, bind, upregulate, resemble, treat, downregulates, palliate, include, and express The first matrix is generated so that the at least some ohmic levels are arranged on each of the horizontal and vertical axes, and the correlation type is displayed at a point where the horizontal and vertical axes intersect, and the graph theory index is the multi-o Including the shortest path between nodes for at least one of the nodes constituting the mixed network, a clustering coefficient for each node, and a centrality coefficient for each node, and the weight of the connection line according to the category of the degree of correlation indicated by the connection line between the nodes Is set differently, the shortest path between nodes is calculated by reflecting the set weight, and the nodes constituting the partial paths are the shortest paths between the nodes for each of the nodes of the multiomics network, clustering for each node A coefficient, and a standard score for at least one of the centrality coefficients for each node is generated by deleting a connection line between a node having a value less than a threshold value and a node having a value less than the threshold value, and the standard score is calculated for each of the nodes of the multiomics network. It is a value obtained by dividing the difference between the index value of the graph theory index and the average index value of the graph theory index for nodes of the multiomics network by a standard error.

The data refiner randomly mixes all the connection lines constituting the multiomics network and then calculates the standard score for each node of the multiomics network, and the number of randomly mixing is 1000 or more.

The data processing apparatus may further include an output unit configured to display some paths having a connection relationship between different ohmic levels in the multiomics network in a hierarchical structure.

The category of the correlation diagram may further include at least one of interaction, cause, present, and localize.

The plurality of ohmic levels may include at least some of a gene level, a protein level, a metabolite level, a symptom level, a disease level, a compound level, a drug level, and a side effect level.

The data generation unit generates a second matrix consisting of the biological entities and the correlation between the biological entities, and connects the biological entities with the correlation between the biological entities, and the second matrix includes the biological entities. They are sequentially arranged on each of the horizontal and vertical axes according to the hierarchical structure of the mix level, and a correlation between the biological entities may be displayed at a point where the horizontal and vertical axes intersect.

In order to execute the data processing method, a recording medium in which a computer-readable program is recorded may be provided.

According to an exemplary embodiment of the present invention, refined information on biological entities related to a predetermined search word and their correlations can be extracted within a short time without searching for an enormous amount of information to discover new drug candidates.

In particular, according to an embodiment of the present invention, it is possible to obtain a multi-omics network composed of only an ohmic level and a correlation degree desired by a user, and a correlation degree between biological entities located at different ohmic levels desired by the user. It can be easily derived, and accordingly, the hierarchical structure of the body and the connection relationship between externally expressed conditions, diseases, and symptoms can be conveniently explored, and disease mechanisms and pharmacological mechanisms can be easily understood.

Accordingly, it is possible to significantly reduce the cost and time required to discover a new drug candidate substance with a high hit rate or a target of a new drug candidate substance.

1 shows the hierarchical structure of the body.
2 illustrates the concept of a network.
3 is a block diagram of a data processing apparatus for discovering a new drug candidate material according to an embodiment of the present invention.
4 is a block diagram of an ohmic DB used by a data processing apparatus for discovering new drug candidate substances according to an embodiment of the present invention.
5 is a flowchart of a data processing method for discovering a new drug candidate substance in a data processing apparatus according to an embodiment of the present invention.
6 shows an example in which an ohmic level is input in step S100 according to an embodiment of the present invention.
7 shows an example in which the type of correlation degree is input in step S110 according to an embodiment of the present invention.
8 shows an example of a first matrix generated in step S130 according to an embodiment of the present invention.
9 shows an example in which a predetermined search word is input.
10 is a part of an example of a second matrix showing a biological entity extracted in step S150 and a degree of correlation therebetween.
11 is an example of a multiomics network created according to an embodiment of the present invention.
12 shows an example of some routes displayed in step S180 according to an embodiment of the present invention.

The present invention is intended to illustrate and describe specific embodiments in the drawings, as various changes may be made and various embodiments may be provided. However, this is not intended to limit the present invention to a specific embodiment, it is to be understood to include all changes, equivalents, and substitutes included in the spirit and scope of the present invention.

Terms including ordinal numbers, such as second and first, may be used to describe various elements, but the elements are not limited by the terms. These terms are used only for the purpose of distinguishing one component from another component. For example, without departing from the scope of the present invention, a second component may be referred to as a first component, and similarly, a first component may be referred to as a second component. The term and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being "connected" or "connected" to another component, it is understood that it may be directly connected or connected to the other component, but other components may exist in the middle. Should be. On the other hand, when a component is referred to as being "directly connected" or "directly connected" to another component, it should be understood that there is no other component in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present application, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof, does not preclude in advance.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms as defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in this application. Does not.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings, but the same reference numerals are assigned to the same or corresponding components regardless of the reference numerals, and redundant descriptions thereof will be omitted.

2 illustrates the concept of a network.

Referring to FIG. 2, a network may be composed of a plurality of nodes, and two nodes may be connected by edges. In the present specification, the network may be a knowledge network, a biological network, or a multiomics network, a node may represent a biological entity, and an edge may represent a degree of correlation between two biological entities.

3 is a block diagram of a data processing apparatus for discovering new drug candidate substances according to an embodiment of the present invention, and FIG. 4 is an ohmic DB used by the data processing apparatus for discovering new drug candidate substances according to an embodiment of the present invention. Is a block diagram of, and FIG. 5 is a flow chart of a data processing method for discovering a new drug candidate substance in a data processing apparatus according to an embodiment of the present invention.

Referring to FIG. 3, the data processing apparatus 100 for discovering a new drug candidate includes a user interface unit 110, a DB extraction unit 120, a data generation unit 130, a data processing unit 140, and a data purification unit ( 150), and an output unit 160 and a storage unit 170.

3 to 5, the user interface unit 110 receives at least some of the ohmic levels among a plurality of levels constituting the ohmic (S100), and at least some of the plurality of types of correlations constituting the ohmic are input. The type of correlation degree is input (S110). Here, omics is also referred to as somatics. For example, there are genetics, transcriptomes, proteomics, metabolomics, epigenomics, and geology, and in detail anatomy, biological pathways, etc. process), a pathway, a pharmacological class, symptoms, diseases, compounds, drugs, side effects, and the like, but are not limited thereto. Multiple ohmic levels are gene level, transcription level, protein level, metabolite level, epigene level, lipid level, anatomical structure level, biological pathway level, conduction pathway level, pharmacological hierarchy level, symptom level, disease level. , Compound level, drug level, side effect level, and the like, but are not limited thereto. Here, the anatomical structure may mean a tissue, an organ, etc., and the biological pathway is a series of cellular components such as location at the level of the intracellular structure, and molecular functions extracted from gene ontology. May be an event of, and the pharmacological layer may be a pharmacological effect, a mechanism of action. And, the plurality of types of interrelationships are "interact", "participate", "covariate", "regulate", "associate", "bind" ", "upregulate", "cause", "resemble", "treat", "downregulates", "palliate", "expression (present)", "localize", "include", and "express" may be included, and identification numbers or identification symbols may be arbitrarily assigned for each type. The identification number or identification symbol for each type may be set by the user or may be automatically set. 6 shows an example in which an ohmic level is input in step S100 according to an embodiment of the present invention, and FIG. 7 shows an example in which a correlation degree type is input in step S110 according to an embodiment of the present invention. Referring to FIG. 6, a screen in which a plurality of ohmic levels can be selected may be exposed through the output unit 160, and at least some of the ohmic levels through the user interface unit 110 among the plurality of ohmic levels Can be chosen. And, referring to FIG. 7, a screen in which a plurality of types of correlations can be selected may be exposed through the output unit 160, and at least a portion of the plurality of types of correlations may be displayed through the user interface unit 110. The type of correlation can be selected.

Next, the DB extracting unit 120 extracts a DB regarding at least some of the ohmic levels selected in step S100 and a DB regarding at least some types of correlations selected in step S110 from the omics database (S120). ). Here, the ohmics DB 200 may be a big data DB, and may be a DB outside the data processing device 100 according to an embodiment of the present invention, and anyone can access or a global network that can be accessed by an authenticated person under predetermined conditions. It can be a public DB. The ohmics DB 200 may pre-store information about an ohmic level and information about a degree of correlation between biological entities within the ohmic level. For example, as shown in FIG. 4, the ohmics DB 200 may include a DB 210 for each ohmic level and a DB 220 for each type of correlation. The DB 210 for each ohmic level is, for example, gene DB, transcription DB, protein DB, metabolite DB, epigene DB, lipid DB, anatomical structure DB, biological pathway DB, conduction pathway DB, symptom DB, It may include disease DB, compound DB, drug DB, and side effect DB. In addition, the DB 220 for each type of correlation is an interaction DB, a participation DB, a covariate DB, a regulate DB, an associate DB, a bind DB, and Upregulate DB, cause DB, resemble DB, treatment DB, downregulates DB, palliate DB, present DB, localize DB , Include DB and express DB. These DBs can be managed and operated by integrating into one big data DB, or distributed and managed and operated.

In addition, the DB extraction unit 120 generates a first matrix consisting of a DB regarding at least some of the ohmic levels extracted in step S120 and a DB regarding at least some types of correlations (S130). Here, the first matrix may be referred to as a set of DBs extracted in step S120. 8 shows an example of a first matrix generated in step S130 according to an embodiment of the present invention. Referring to FIG. 8, ohmic levels selected in step S100 are disposed on each of the horizontal and vertical axes, and the types of correlations selected in step S110 may be generated to be displayed at points where the horizontal and vertical axes intersect. For example, gene level, protein level, metabolic level, anatomical structure level, conduction pathway level, biological pathway level, compound level, side effect level, disease level, pharmacological stratification level and symptom level are the horizontal axis of the first matrix. And it can be arranged on each of the vertical axis, the interaction (interact, Int), participation (participate, P), covariate (Co), regulation (regulate, Reg), which are types of correlation at the point where the horizontal axis and the vertical axis intersect. , Associate (A), bind (B), upregulate (U), cause (ca), resemble (R), treatment (T), downregulates , D), palliate (Pa), expression (present, Pr), location (localize, L), include (Inc), and at least one of expression (express, E) may be displayed.

Meanwhile, the user interface unit 110 receives a predetermined search word (S140). The predetermined search word may be a search word that the user wants to search for information, and one of a plurality of biological entities included in each ohmic level, for example, a gene name, a protein name, a metabolite name, a symptom name, a disease name, a compound It may include one of name, drug name, or side effect name. 9 shows an example in which a predetermined search word is input. Referring to FIG. 9, a screen for inputting a predetermined search word may be exposed through the output unit 160, and a predetermined search word may be input through the user interface unit 110. 9 shows an example of selecting a disease name as a category and inputting epilepsy syndrome as a predetermined search word.

Next, the data generation unit 130 extracts at least one biological entity related to the predetermined search word received in step S140 using the first matrix generated in step S130, and uses the first matrix generated in step S130. Thus, the correlation between the predetermined search word and the extracted biological entity is extracted (S150). Here, the biological entity may include at least one of genes, proteins, metabolites, symptoms, diseases, compounds, and drugs, and the ohmic level to which the predetermined search word belongs may be the same as the ohmic level to which the biological entity belongs. , It may be different. For example, as illustrated in FIG. 9, when a predetermined search word is epilepsy syndrome, which is a disease name, the biological entity extracted in step S150 is a gene associated with epilepsy syndrome, a protein associated with epilepsy syndrome, and a metabolite associated with epilepsy syndrome. , symptoms associated with epilepsy syndrome, diseases associated with epilepsy syndrome, compounds associated with epilepsy syndrome, and medications associated with epilepsy syndrome. To this end, the data generation unit 130 comprises a gene DB, a protein DB, a metabolite DB, an anatomical structure DB, a conduction path DB, a biological path DB, a compound DB, a side effect DB, a disease constituting the first matrix in step S130. Biological entities related to epilepsy syndrome can be extracted from each of the DB, pharmacological layer DB, and symptom DB. Accordingly, the biological entities extracted in step S150 are multiple genes associated with epilepsy syndrome, multiple proteins associated with epilepsy syndrome, multiple metabolites associated with epilepsy syndrome, multiple symptoms associated with epilepsy syndrome, multiple symptoms associated with epilepsy syndrome. It may include at least one of a plurality of compounds associated with the disease, epilepsy syndrome, and a plurality of drugs associated with epilepsy syndrome.

In this way, when extracting biological entities and correlations associated with a predetermined search word using the first matrix of step S130, the amount of DB to be searched can be significantly reduced, and accordingly, time and cost for searching information Can be reduced, and it is possible to extract only the information the user wants.

In this case, in order for the data generating unit 130 to extract a correlation between at least one biological entity and biological entities related to a predetermined search word, the data generating unit 130 is based on artificial intelligence technology including machine learning, Natural language processing algorithms can be used. Here, natural language processing refers to various technologies that mechanically analyze language phenomena spoken by humans to make them into a form that can be understood by a computer, and express the form that can be understood by a computer in a language that can be understood by humans. To this end, the ohmics DB 200 may be a language-based DB for each biological entity type, and may include information reflecting a machine learning result and a feedback result.

Alternatively, in order for the data generating unit 130 to extract a correlation between at least one biological entity and biological entities related to a predetermined search word, the data generating unit 130 is based on artificial intelligence technology including machine learning, Deep neural network algorithms can also be used. Here, the deep neural network is an artificial neural network consisting of several hidden layers between the input layer and the output layer, and refers to all technologies used for classification, prediction, image recognition, and character recognition. To this end, the ohmics DB 200 may be an image-based DB for each biological entity type, and may include information reflecting a machine learning result and a feedback result.

10 is a part of an example of a second matrix showing a biological entity extracted in step S150 and a degree of correlation therebetween. Referring to FIG. 10, in the second matrix, a plurality of biological entities are sequentially disposed on each of the horizontal and vertical axes according to the hierarchical structure of the ohmic level, and the correlation between the plurality of biological entities is at a point where the horizontal and vertical axes intersect. It can be created in the way it is displayed. For example, the ohmic level selected in step S100 is a gene level, a conduction pathway level, a protein level, a metabolite level, a disease level, a side effect level, and a compound level, and the predetermined search word input in step S140 is one of the compounds. In the case of bupropion, in step S150, a plurality of genes, a plurality of conduction pathways, a plurality of proteins, a plurality of metabolites, a plurality of diseases, and a plurality of It can be seen that side effects, a plurality of compounds, are extracted as biological entities, and these biological entities are sequentially arranged on the horizontal axis and the vertical axis according to the hierarchical structure of the ohmic level. In addition, it can be seen that the correlation between biological entities is displayed in different colors at the point where the horizontal axis and the vertical axis intersect.

The shape of the second matrix is exemplary, and is not limited thereto, and may be modified in various shapes.

Next, the data generation unit 130 generates a multiomics network by using the result extracted in step S150 (S160). 11 is an example of a multiomics network created according to an embodiment of the present invention. Here, the multiomics network uses a predetermined search word received in step S140 and biological entities extracted in step S150 as nodes, and a degree of correlation between a predetermined search word extracted in step S150 and biological entities or a degree of correlation between biological entities According to the present invention, a plurality of nodes may be connected using a connection line. Paths from Node A, which is one of the nodes in the multiomics network, to Node B, which is the other, may vary, and all possible paths may be connected by connection lines. Here, the multiomics network is a network consisting of a degree of interrelationship between biological entities and may be mixed with a biological network. In the multiomics network, some of the plurality of biological entities that become nodes may be included in different levels of omics from the remaining biological entities. That is, as illustrated in FIG. 11, the multiomics network includes a plurality of different ohmic levels such as gene level, conduction pathway level, protein level, metabolite level, compound level, side effect level, and disease level. The biological entity is a node, and some of the plurality of biological entities included in the gene level may be connected to some of the plurality of biological entities included in the protein level or may be connected to some of the plurality of biological entities included in the conduction pathway level. Likewise, some of the plurality of biological entities included in the compound level are linked to some of the plurality of biological entities included in the protein level, some of the plurality of biological entities included in the conduction pathway level, or included in the side effect level. It may be connected to some of the plurality of biological entities.

As described above, according to an embodiment of the present invention, when some of the plurality of ohmic levels and some of the types of the plurality of correlations are input through the user interface unit 110, the DB and the corresponding ohmic level are Since the DB related to the type of correlation is automatically extracted, the amount of information to be searched by the data processing device 100 can be significantly reduced, and accordingly, a multi-omics network composed of the desired ohmic level and the correlation type Can be obtained. In addition, according to an embodiment of the present invention, when some of the plurality of ohmic levels and some of the plurality of types of correlation are input through the user interface unit 110, the user's desired ohmic level and the type of correlation are received. A multi-omics network composed of can be obtained, and accordingly, a hierarchical structure between a plurality of biological entities associated with a predetermined search word within a desired ohmic level of a user can be easily grasped.

Next, the data refiner 150 extracts some paths in the multiomics network generated in step S160 (S170), and the output unit 160 displays some paths extracted in step S170 (S180). Here, some paths may be paths that connect some nodes extracted from biological entities in the multiomics network, which are relatively high related to a predetermined search word among a plurality of paths that connect biological entities in the multiomics network. It may be a path determined to be of high importance or a path determined to be of relatively high importance. 12 shows an example of some routes displayed in step S180 according to an embodiment of the present invention. Referring to FIG. 12, some paths having a connection relationship between different ohmic levels in the multiomics may be displayed in a hierarchical structure. Accordingly, it is possible to intuitively grasp the degree of correlation between a plurality of biological entities included in different ohmic levels.

Meanwhile, in order to extract some paths in the multi-omics network as in step S170, the data processing unit 140 may generate a graph theory index of the multi-omics network, and the data refiner 150 is the data processing unit 140 In terms of the graph theory index generated in, some nodes in the multiomics network with high correlation can be extracted.

To this end, the graph theory index may include at least one of a shortest path between nodes for a plurality of nodes constituting a multiomics network, a clustering coefficient for each node, a centrality coefficient for each node, and a hub characteristic for each node.

The shortest path between nodes may mean the shortest path among numerous paths from node A to node B in a multiomics network. Hereinafter, a method of calculating the shortest path between Node A as one of the biological entities and Node B as the other one of the biological entities will be described.

There are various paths from node A to node B, and node A and node B may be directly connected, or at least one intermediate node may exist on each path between node A and node B.

The shortest path between node A and node B can be obtained using the number of intermediate nodes per path. For example, among various paths between node A and node B, as the number of intermediate nodes decreases, it may be determined that the path is shorter.

Alternatively, the shortest path between node A and node B is obtained using the number of intermediate nodes for each path, but may reflect the type of interrelationship for each connection line. That is, weights are set differently for each category of correlation, and weights may be applied to correlations that exist for each path.

Equation 1 is an example of an equation for calculating the shortest path between nodes.

는 두 노드 i와 j 사이의 최단 경로이다. 경로 별로 수학식 1의 값을 구하며, 가장 낮은 값 또는 가장 높은 값을 가지는 경로가 최단 경로로 선택될 수 있다. Where w _st is an index of correlation between two nodes s and t, f is a weight transformation function,

Is the shortest path between two nodes i and j. A value of Equation 1 is obtained for each path, and a path having the lowest value or the highest value may be selected as the shortest path.

Next, a clustering coefficient for each node may be calculated by Equation 2 and Equation 3. Here, the clustering coefficient may be referred to as a grouping coefficient, and may mean a probability that a specific node and neighboring nodes are connected to each other or a connection density between a specific node and neighboring nodes.

Here, t ^w _i means the number of triangles in the graph created around each node i of the multiomics network, N is the total node set of the _multiomics network, and w _ij is the correlation between two nodes i and j. Is an index, w _ih is a correlation index between two nodes i and h, and w _jh is a correlation index between two nodes j and h.

Here, C ^w means the clustering coefficient, t ^w _i is the number of triangles in the graph created around each node i of the multiomics network, and k _i is the degree of node i, that is, the multiomics network of node i. It means the value of my degree of connectivity.

Next, the centrality index for each node is an index for whether a specific node has the function of a _{hub.D nodal} (nodal degree) value, BC (betweenness centrality) value, E _nodal (nodal efficiency) value, etc. Can be represented by Here, the value of D _{nodal is a} value of the degree of connectivity within the multiomics network of each node, that is, an index indicating how strong or weak node i has connectivity in the _{multiomics network, and the} value of E _nodal is the knowledge of node i. A value of the degree of efficiency within the network, that is, a value expressed as the reciprocal of the shortest path in Equation 1, and the shorter the path, the higher the efficiency. It is an indicator.

First, the value of D _nodal can be calculated by Equation 4.

Here, w _ij is a correlation index between two nodes i and j, and N is a set of all nodes of the multiomics network.

And, the E _nodal value may be calculated by Equation 5.

는 수학식 1에서 계산한 최단 경로를 나타내는 값이다.Here, N is the set of all nodes of the multiomics network,

Is a value representing the shortest path calculated in Equation 1.

Next, Betweenness centrality (BC) can be calculated by Equation 6.

는 노드 h 와 j 사이의 최단 거리를 의미하고,

는 노드 i를 통과하는 h 와 j 사이의 최단 거리를 의미한다. here,

Is the shortest distance between nodes h and j ,

Denotes the shortest distance between h and j passing through node i .

Next, when it is determined that the predetermined node has the function of the hub, the characteristics of the hub are classified. At this time, the nature of the hub can be classified into a kinless hub, a connector hub, and a provincial hub. Here, a kinless hub refers to a hub with the highest influence, that is, a hub connected to nodes in many modules, a connector hub refers to a hub that connects modules within a multiomics network, and a provincial hub is mainly within a module. It means a hub with high influence. Here, the module may be a structural configuration group in which the entire multiomics network is subdivided.

To this end, the module index (Modularity) in the multiomics network may be calculated as in Equation 7. The module index (modularity) refers to the number of module types in the entire multiomics network.

는 노드 i에서의 가중치 합을 의미하고,

는 가중치 합을 의미한다. δ_mi,mj는 크로네커의 델타(kronecker delta)이고, mi=mj인 경우 1이고, 나머지인 경우 0이다. here,

Means the sum of weights at node i ,

Means the sum of weights. δ _mi,mj is the kronecker delta, 1 when mi=mj, and 0 when the remainder.

Next, the participation coefficient (PC) of the multiomics network module may be calculated as shown in Equation 8.

는 모듈 m 내에서 노드 i와 나머지 모든 노드 간의 연결 수를 의미하고, 모듈 m은 전체 멀티오믹스 네트워크를 세분화한 구조적 구성 그룹을 의미한다.Here, M means a set of modules,

Is in the module m means the node i and the number of connections between all other nodes, modules and m is the mean structural group a granular mix the entire multi-O network.

In addition, a z score (within-module degree) of the multiomics network module may be calculated as in Equation 9.

는 노드 i의 모듈 m 내에서의 연결 정도(degree)를 의미하며,

는 각각 모듈 m내의 연결 정도 분포(degree distribution)의 평균과 표준 편차를 의미한다.Here, m and _i denotes the node i in the module m,

Means the degree of connection in module m of node i ,

Denotes the mean and standard deviation of the degree distribution in module m , respectively.

It is possible to distinguish whether each node is a hub in the module through the calculation of the index of Equation 9 above. For example, as follows, when the Z score of the multiomics network module is 2.5 or higher, it may be determined as a hub.

1.within-module z-score = 2.5: hub

2. within-module z-score <2.5: not hub

In addition, when it is determined that the node is a hub in the module, the type of the hub can be classified as follows through the calculation of the index of Equation 8, and an example of the type of the hub according to the PC is as follows.

1.Provincial Hub: PC =0.30

2. Connector Hub: 0.3 <PC =0.75

3. Kinless Hub: PC> 0.75

At this time, the nodes constituting some paths extracted in step S170 are the index values for the shortest path between nodes among the plurality of nodes constituting the multiomics network in step S160, the index values for the clustering coefficient for each node, and the centrality for each node. At least some of the index values for the coefficient may be some nodes having a threshold value or more. That is, some of the paths extracted in step S170 are at least one of an index value for the shortest path between nodes, an index value for a clustering coefficient for each node, and an index value for a centrality coefficient for each node among a plurality of nodes constituting the multiomics network. It may be created by deleting a node whose part is less than the threshold value and deleting a connection associated with the deleted node.

Here, the graph theory index compared with the threshold value may be an index value for a shortest path between nodes, an index value for a clustering coefficient for each node, and an index value for a centrality coefficient for each node. Alternatively, the graph theory index compared to the threshold value may be a value calculated by integrating at least two of an index value for a shortest path between nodes, an index value for a clustering coefficient for each node, and an index value for a centrality coefficient for each node. .

At this time, at least one of the indicator value for the shortest path between nodes, the indicator value for the clustering coefficient for each node, and the indicator value for the centrality coefficient for each node may be calculated as a standard score for each node, and the calculated standard score is a threshold value. Can be compared to

Here, the standard score may be a z score, and the threshold value may mean 95% significance.

The Z score can be calculated as in Equation 10.

로 나타낼 수 있으며, σ는 표준 편차이고, n은 멀티오믹스 네트워크를 구성하는 복수의 노드의 개수이다. Here, z is a z score, X is an index value of a predetermined graph theory index for a specific node in a multiomics network, and mean(x) is a predetermined graph theory index for a plurality of nodes in a multiomics network. It is the average index value, and SE(x) is the standard error of the index value of a given graph theory index in the multiomics network. here,

It can be expressed as, σ is the standard deviation, and n is the number of nodes constituting the multiomics network.

That is, the z-score is the standard difference between the index value of a predetermined graph theory index for each node constituting the multiomics network and the average index value of a predetermined graph theory index for a plurality of nodes constituting the multiomics network. It can be a value divided by the error.

In this case, the z score may be calculated through a permutation test. The permutation test may be performed by randomly mixing all the connecting lines constituting the multiomics network and then calculating a z score for each node. At this time, the number of random mixing may be 1000 or more.

Alternatively, the nodes constituting some paths extracted in step S170 may be some nodes extracted by using an index value for the hub characteristic for each node from among a plurality of nodes constituting the multiomics network. That is, the node constituting some paths extracted in step S170 is a node determined to be a hub within the module through the index calculation of Equation 9, preferably a node classified as one of a kinless hub, a connector hub, and a provincial hub, more preferably For example, it may be a node classified as one of a kinless hub and a connector hub, more preferably a node classified as a kinless hub.

Meanwhile, the data processing apparatus 100 according to an embodiment of the present invention may include a data storage unit 170. The data storage unit 170 may be connected to the data generation unit 130, the data processing unit 140 and the data purification unit 150, and the data generation unit 130, the data processing unit 140, and the data purification unit 150 You can save the result calculated from. The data storage unit 170 may be connected wirelessly or wired to an external learning server, and may transmit stored data to an external learning server.

The term'~ unit' used in this embodiment refers to software or hardware components such as field-programmable gate array (FPGA) or ASIC, and'~ unit' performs certain roles. However,'~ part' is not limited to software or hardware. The'~ unit' may be configured to be in an addressable storage medium or may be configured to reproduce one or more processors. Thus, as an example,'~ unit' refers to components such as software components, object-oriented software components, class components and task components, processes, functions, properties, and procedures. , Subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, database, data structures, tables, arrays, and variables. The components and functions provided in the'~ units' may be combined into a smaller number of elements and'~ units', or may be further divided into additional elements and'~ units'. In addition, components and'~ units' may be implemented to play one or more CPUs in a device or a security multimedia card.

Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art will variously modify and change the present invention within the scope not departing from the spirit and scope of the present invention described in the following claims. You will understand that you can do it.

Claims (13)

In the data processing method for discovering new drug candidate substances performed in a data processing device,
Receiving at least some of the ohmics levels forming the ohmics through a user interface;
Receiving at least some types of correlations among the plurality of types of correlations forming the ohmic through a user interface;
Selecting a DB for the at least some of the ohmic levels and a DB for the at least some types of correlations from an ohmics DB including data for each ohmic level and data for each type of correlation;
Generating a first matrix consisting of a DB related to the at least part of the ohmic level and a DB related to the type of the at least part of the correlation;
Receiving a search word input through a user interface;
Extracting biological entities related to the search word and belonging to an ohmic level different from that of the search word from the first matrix, and extracting a degree of correlation between the search word and the biological entities;
Generating a multiomics network in which nodes representing the search word and the biological entities are connected according to a degree of correlation between the search word and the biological entities or a degree of correlation between the biological entities;
Generating a graph theory index for each of the nodes of the multiomics network; And
Including the step of extracting some paths having a connection relationship between different ohmic levels in the multiomics network by using some of the nodes extracted using the graph theory index,
Some of the biological entities are included in different ohmic levels from the other biological entities,
The search word includes at least one of a gene name, a protein name, a metabolite name, a symptom name, a disease name, a compound name, and a drug name,
The biological entity includes at least one of genes, proteins, metabolites, symptoms, diseases, compounds and drugs,
The categories of correlation are participate, covariate, regulate, associate, bind, upregulate, resemble, treat, and down. Includes downregulates, palliate, include, and express,
The first matrix is generated such that the at least some ohmic levels are disposed on each of a horizontal axis and a vertical axis, and the type of correlation is displayed at a point where the horizontal axis and the vertical axis intersect,
The graph theory indicator includes a shortest path between nodes for at least one of nodes constituting the multiomics network, a clustering coefficient for each node, and a centrality coefficient for each node,
The weight of the connection line is set differently according to the category of the degree of correlation indicated by the connection line between the nodes, and the shortest path between the nodes is calculated by reflecting the set weight,
Nodes constituting the some paths,
A connection line between a node having a standard score for at least one of the shortest path between nodes for each of the nodes of the multiomics network, a clustering coefficient for each node, and a centrality coefficient for each node is less than a threshold value and a node less than the threshold value Is created by deleting
The standard score is data obtained by dividing the difference between the index value of the graph theory index for each node of the multiomics network and the average index value of the graph theory index for nodes of the multiomics network by a standard error Processing method. The method of claim 1,
The step of extracting the partial path,
Randomly mixing all the connecting lines constituting the multiomics network, and then calculating the standard score for each of the nodes of the multiomics network,
The number of random mixing is 1000 or more data processing method. The method of claim 1,
And displaying some paths having a connection relationship between different ohmic levels in the multiomics network in a hierarchical structure. The method of claim 1,
The category of the correlation diagram further includes at least one of interaction, cause, present, and localize. The method of claim 1,
The plurality of ohmic levels include at least some of a gene level, a protein level, a metabolite level, a symptom level, a disease level, a compound level, a drug level, and a side effect level. The method of claim 1,
Generating the multiomics network,
Generating a second matrix consisting of the biological entities and a degree of correlation between the biological entities; And
Including the step of connecting the biological entities to the relationship between the biological entities,
In the second matrix, the biological entities are sequentially arranged on a horizontal axis and a vertical axis according to a hierarchical structure of an ohmic level, and a correlation between the biological entities is displayed at a point where the horizontal axis and the vertical axis intersect. In the data processing device for discovering new drug candidate substances,
A user interface unit that receives at least some of the ohmic levels from among the plurality of ohmics levels that make up the ohmics, and receives at least some of the types of correlations that make up the ohmics;
From an ohmics DB including data for each ohmic level and data for each type of correlation, selects the at least partial DB for the ohmic level and the DB for the at least partial correlation type, and A DB extracting unit that generates a first matrix consisting of a DB related to a mix level and a DB related to a type of the at least some correlations;
Extracting biological entities related to the search word and belonging to an ohmic level different from the ohmic level of the search word input through the user interface unit from the first matrix, extracting a degree of correlation between the search word and the biological entities, A data generator configured to generate a multiomics network in which a plurality of nodes representing the search word and the biological entities are connected according to the correlation between the search word and the biological entities or the correlation between the biological entities;
A data processing unit generating a graph theory index for each of the nodes of the multiomics network; And
A data refiner for extracting some paths having a connection relationship between different ohmic levels in the multiomics network by using some nodes extracted using the graph theory index among the nodes,
Some of the biological entities are included in different ohmic levels from the other biological entities,
The search word includes at least one of a gene name, a protein name, a metabolite name, a symptom name, a disease name, a compound name, and a drug name,
The biological entity includes at least one of genes, proteins, metabolites, symptoms, diseases, compounds and drugs,
The categories of correlation are participate, covariate, regulate, associate, bind, upregulate, resemble, treat, and down. Include downregulates, palliate, include and express,
The first matrix is generated such that the at least some ohmic levels are disposed on each of a horizontal axis and a vertical axis, and the type of correlation is displayed at a point where the horizontal axis and the vertical axis intersect,
The graph theory indicator includes a shortest path between nodes for at least one of nodes constituting the multiomics network, a clustering coefficient for each node, and a centrality coefficient for each node,
The weight of the connection line is set differently according to the category of the degree of correlation indicated by the connection line between the nodes, and the shortest path between the nodes is calculated by reflecting the set weight,
Nodes constituting the some paths,
A connection line between a node having a standard score for at least one of the shortest path between nodes for each of the nodes of the multiomics network, a clustering coefficient for each node, and a centrality coefficient for each node is less than a threshold value and a node less than the threshold value Is created by deleting
The standard score is data obtained by dividing the difference between the index value of the graph theory index for each node of the multiomics network and the average index value of the graph theory index for nodes of the multiomics network by a standard error Processing device. The method of claim 7,
The data refinement unit,
After randomly mixing all the connecting lines constituting the multiomics network, the standard score is calculated for each of the nodes of the multiomics network,
The number of random mixing is 1000 or more data processing device. The method of claim 7,
The data processing apparatus further comprises an output unit configured to display some paths having a connection relationship between different ohmic levels in the multiomics network in a hierarchical structure. The method of claim 7,
The category of the correlation diagram further includes at least one of interaction, cause, present, and localize. The method of claim 7,
The plurality of ohmic levels includes at least some of a gene level, a protein level, a metabolite level, a symptom level, a disease level, a compound level, a drug level, and a side effect level. The method of claim 7,
The data generation unit generates a second matrix consisting of the biological entities and a degree of correlation between the biological entities, and connects the biological entities with a degree of correlation between the biological entities,
In the second matrix, the biological entities are sequentially arranged on a horizontal axis and a vertical axis according to a hierarchical structure of an ohmic level, and a correlation between the biological entities is displayed at a point where the horizontal axis and the vertical axis intersect. A recording medium in which a computer-readable program is recorded for executing the data processing method performed in any one of claims 1 to 6.

Images