KR20240059070A - Method of prediction based on graph neural network - Google Patents

Abstract

According to an embodiment of the present disclosure, a method for performing prediction based on a graph neural network model is disclosed. Specifically, according to the present disclosure, a computing device inputs information into a graph neural network model, utilizes the graph neural network model, and performs prediction based on the input information, wherein the graph neural network model includes classification of nodes and It corresponds to a model learned using a loss function calculated based on the frequency of occurrence of each of the plurality of related classes and the distance information associated with the plurality of classes.

Description

Method for making predictions based on a graph neural network model {METHOD OF PREDICTION BASED ON GRAPH NEURAL NETWORK}

The present invention relates to a method of performing prediction based on a graph neural network model, and specifically, utilizes a model learned based on the frequency of occurrence of each of a plurality of classes related to the classification of a node and the distance information associated with the plurality of classes. Thus, it relates to a method of performing prediction based on input information.

A Graph Neural Network is an artificial neural network model that receives inputs such as the graph structure as information and can perform well the task of classifying the nodes of each graph. In node classification problems using graph neural networks, the cross entropy loss function is mainly used to learn the graph artificial neural network. The cross-entropy loss function is a loss function that reflects the predicted probability that each node will be classified into an individual class as a log function, and as the predicted probability converges to the actual label value, the value of the loss function approaches 0.

In a specific node classification problem, there is a need to reflect the frequency of occurrence of each class into which the node is classified or the topological distance information between nodes included in the classified classes. In solving this type of node classification problem, if a graph neural network model is trained using only the cross-entropy loss function, there may be a limit to improving the model's performance because the predictions of individual nodes are unrelated and independent. For example, there are cases where the nodes classified in the inference step are scattered on the graph, or a specific class occurs repeatedly and does not function as a possible answer to the problem.

Therefore, we develop a loss function that can reflect the frequency of occurrence of each class or the topological distance information of nodes included in the classified class, and use this to perform predictions based on the learned graph neural network model. The industry demand exists.

The present disclosure was developed in response to the above-described background technology, and in the node classification problem of a graph neural network, it utilizes a model learned based on the frequency of occurrence and distance information of each of a plurality of classes related to the classification of nodes, and inputs The purpose is to make predictions based on information.

Meanwhile, the technical problem to be achieved by the present disclosure is not limited to the technical problems mentioned above, and may include various technical problems within the scope of what is apparent to those skilled in the art from the contents described below.

According to an embodiment of the present disclosure for realizing the above-described problem, a method of performing prediction based on a graph neural network model is disclosed. The method includes inputting information into a graph neural network model; And performing a prediction based on the input information using the graph neural network model, wherein the graph neural network model includes the frequency of occurrence of each of a plurality of classes related to the classification of the node, and It can correspond to a learned model using a loss function calculated based on distance information associated with a plurality of classes.

In an alternative embodiment, the prediction includes a prediction related to the binding potential of the compound with the kinase, and the node corresponds to an atom contained in the compound.

In an alternative embodiment, the plurality of classes may be associated with a plurality of binding sites of the kinase.

In an alternative embodiment, the distance information associated with the plurality of classes may include topological distance information between nodes associated with the plurality of classes.

In an alternative embodiment, the loss function includes a first loss calculated based on the norm between the class information matrix associated with the label value of the node included in the training data and the class information matrix associated with the predicted value of the node. Can contain functions.

In an alternative embodiment, the class information matrix associated with the label value of the node is calculated based on a label value matrix including label values of a plurality of nodes and a distance matrix including distance information between the plurality of nodes, and , the class information matrix related to the prediction value of the node may be calculated based on a prediction value matrix including prediction values of the plurality of nodes and a distance matrix including distance information between the plurality of nodes.

In an alternative embodiment, each class information matrix includes: a diagonal element containing information about the frequency of occurrence of each class; and an off-diagonal element containing topological distance information between nodes associated with different classes.

In an alternative embodiment, the loss function may further include a second loss function calculated based on cross entropy of the training data.

In an alternative embodiment, the loss function may be determined as a weighted sum operation of the first loss function and the second loss function.

According to an embodiment of the present disclosure for realizing the above-described task, a computer program for performing prediction based on a graph neural network model is disclosed. When executed by one or more processors, the computer program causes the one or more processors to perform operations for inserting punctuation marks, the operations including: inputting information into the graph neural network model; And an operation of performing prediction based on the input information using the graph neural network model, wherein the graph neural network model includes the frequency of occurrence of each of a plurality of classes related to the classification of the node, and the plurality of classes. It can be matched to a learned model using a loss function calculated based on the distance information associated with the data.

According to an embodiment of the present disclosure for realizing the above-described problem, a computing device for performing prediction based on a graph neural network model is disclosed. The computing device includes at least one processor; and memory; wherein the at least one processor inputs information into the graph neural network model and performs prediction based on the input information using the graph neural network model, wherein the graph neural network model classifies nodes. It may correspond to a model learned using a loss function calculated based on the frequency of occurrence of each of the plurality of classes related to and the distance information associated with the plurality of classes.

According to an embodiment of the present disclosure for realizing the above-described task, a method for training a graph neural network model is disclosed. The learning method of the graph neural network includes calculating a first class information matrix related to label values of a plurality of nodes; calculating a second class information matrix related to predicted values of the plurality of nodes; calculating a loss function based on the first class information matrix and the second class information matrix; and training the graph neural network model based on the loss function, wherein each of the first class information matrix and the second class information matrix is an occurrence frequency of each of a plurality of classes, and the plurality of classes. It may include distance information associated with .

In an alternative embodiment, calculating the first class information matrix includes a label value matrix including label values of the plurality of nodes, and a distance matrix including distance information between the plurality of nodes. Comprising a step of calculating a first class information matrix, and calculating a second class information matrix, calculating second class information based on a prediction value matrix including prediction values of the plurality of nodes, and the distance matrix. It may include a step of calculating a matrix.

The present disclosure can provide a graph neural network model with improved performance. For example, the present disclosure may provide a graph neural network model that considers the frequency of occurrence of each of a plurality of classes related to the classification of a node and distance information (eg, topological distance information) associated with the plurality of classes.

The following drawings attached to be used in explaining the embodiments of the present disclosure are only some of the embodiments of the present disclosure, and for those skilled in the art (hereinafter referred to as “ordinary skilled in the art”), Other drawings can be obtained based on these drawings without effort leading to new inventions.
1 is a block diagram of a computing device for inserting punctuation marks according to an embodiment of the present disclosure.
Figure 2 is a schematic diagram showing a network function according to an embodiment of the present disclosure.
Figure 3 is a flowchart showing a process of performing prediction based on a graph neural network model according to an embodiment of the present disclosure.
Figure 4 is a flowchart showing a method of learning a graph neural network model according to an embodiment of the present disclosure.
Figure 5 is a conceptual diagram showing predicting the position of a hydrogen bond atom of a compound that binds to a kinase hinge according to an embodiment of the present disclosure.
FIG. 6 is a conceptual diagram illustrating a label value matrix, a prediction value matrix, and distance information between a plurality of nodes according to an embodiment of the present disclosure.
7 is a brief, general schematic diagram of an example computing environment in which embodiments of the present disclosure may be implemented.

The present disclosure trains a graph neural network model using a loss function calculated based on the frequency of occurrence of each of a plurality of classes related to the classification of a node and the distance information associated with the plurality of classes, and uses the graph neural network model to learn the input A method of performing prediction based on information is disclosed.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are presented to provide an understanding of the disclosure. However, it is clear that these embodiments may be practiced without these specific descriptions.

As used herein, the terms “component,” “module,” “system,” and the like refer to a computer-related entity, hardware, firmware, software, a combination of software and hardware, or an implementation of software. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, a thread of execution, a program, and/or a computer. For example, both an application running on a computing device and the computing device can be a component. One or more components may reside within a processor and/or thread of execution. A component may be localized within one computer. A component may be distributed between two or more computers. Additionally, these components can execute from various computer-readable media having various data structures stored thereon. Components may transmit signals, for example, with one or more data packets (e.g., data and/or signals from one component interacting with other components in a local system, a distributed system, to other systems and over a network such as the Internet). Depending on the data being transmitted, they may communicate through local and/or remote processes.

Additionally, the term “or” is intended to mean an inclusive “or” and not an exclusive “or.” That is, unless otherwise specified or clear from context, “X utilizes A or B” is intended to mean one of the natural implicit substitutions. That is, either X uses A; X uses B; Or, if X uses both A and B, “X uses A or B” can apply to either of these cases. Additionally, the term “and/or” as used herein should be understood to refer to and include all possible combinations of one or more of the related listed items.

Additionally, the terms “comprise” and/or “comprising” should be understood to mean that the corresponding feature and/or element is present. However, the terms “comprise” and/or “comprising” should be understood as not excluding the presence or addition of one or more other features, elements and/or groups thereof. Additionally, unless otherwise specified or the context is clear to indicate a singular form, the singular terms herein and in the claims should generally be construed to mean “one or more.”

And, the term “at least one of A or B” should be interpreted to mean “when it contains only A,” “when it contains only B,” or “when it is a combination of A and B.”

Those skilled in the art will additionally recognize that the various illustrative logical blocks, components, modules, circuits, means, logic, and algorithm steps described in connection with the embodiments disclosed herein may be implemented using electronic hardware, computer software, or a combination of both. It must be recognized that it can be implemented with To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented in hardware or software will depend on the specific application and design constraints imposed on the overall system. A skilled technician can implement the described functionality in a variety of ways for each specific application. However, such implementation decisions should not be construed as causing a departure from the scope of the present disclosure.

The description of the presented embodiments is provided to enable anyone skilled in the art to use or practice the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Therefore, the present disclosure is not limited to the embodiments presented herein. This disclosure is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

In this disclosure, network function, artificial neural network, and neural network may be used interchangeably.

1 is a block diagram of a computing device for performing prediction based on a graph neural network model according to an embodiment of the present disclosure.

The configuration of the computing device 100 shown in FIG. 1 is only a simplified example. In one embodiment of the present disclosure, the computing device 100 may include different components for performing the computing environment of the computing device 100, and only some of the disclosed components may configure the computing device 100.

The computing device 100 may include a processor 110, a memory 130, and a network unit 150.

The processor 110 may be composed of one or more cores, and may include a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU) of a computing device. unit) may include a processor for data analysis and deep learning. The processor 110 may read a computer program stored in the memory 130 and perform data processing for machine learning according to an embodiment of the present disclosure. According to an embodiment of the present disclosure, the processor 110 may perform an operation for learning a neural network. The processor 110 is used for learning neural networks, such as processing input data for learning in deep learning (DL), extracting features from input data, calculating errors, and updating the weights of the neural network using backpropagation. Calculations can be performed.

At least one of the CPU, GPGPU, and TPU of the processor 110 may process learning of the network function. For example, CPU and GPGPU can work together to process learning of network functions and data classification using network functions. Additionally, in one embodiment of the present disclosure, the processors of a plurality of computing devices can be used together to process learning of network functions and data classification using network functions. Additionally, a computer program executed in a computing device according to an embodiment of the present disclosure may be a CPU, GPGPU, or TPU executable program.

According to an embodiment of the present disclosure, the processor 110 utilizes a loss function calculated based on the graph neural network model, the frequency of occurrence of each of a plurality of classes related to the classification of the node, and the distance information associated with the plurality of classes. You can learn it and perform predictions based on the input information using a graph neural network model.

For example, predictions of a graph neural network may include predictions related to the likelihood of a compound binding to a kinase. In this case, all atoms included in the compound can become nodes, or among the atoms included in the compound, only N (nitrogen) atoms and O (oxygen) atoms that can form hydrogen bonds can become nodes. Additionally, a class may be associated with a plurality of binding positions present in a kinase, and distance information associated with the classes may be associated with topological distance information between nodes associated with the classes. However, the prediction of the graph neural network of the present disclosure is not limited to the example of the possibility of combination of a kinase and a compound described above. A specific method by which the processor 110 trains the graph neural network and performs prediction will be described later with reference to FIG. 3.

For prediction of a graph neural network, the processor 110 generates a first loss function based on the norm between the class information matrix related to the label value of the node included in the learning data and the class information matrix related to the predicted value of the node. can be calculated. At this time, the class information matrix related to the label value for calculating the loss function may be calculated based on the label value matrix including the label values of a plurality of nodes and the distance matrix including distance information between the plurality of nodes. . Additionally, a class information matrix related to the predicted value of a node may be calculated based on a predicted value matrix including predicted values of a plurality of nodes and a distance matrix including distance information between a plurality of nodes. The specific process of calculating the first loss function will be described later with reference to FIG. 3.

The processor 110 may calculate a loss function and train a graph neural network model based on the loss function, then input data to the graph neural network and perform prediction. For example, the input data may include data related to the sequence information of the kinase, the structural information of the kinase, or the structural information of the compound that binds to the kinase, but the present disclosure is not limited thereto.

According to one embodiment of the present disclosure, in order to train a graph neural network model, the processor 110 calculates a loss based on a first class information matrix related to the label values of the nodes and a second class information matrix related to the predicted values of the nodes. Functions can be computed. The processor may then train a graph neural network model based on the computed loss function. A method of training a graph neural network model according to the present disclosure will be described later with reference to FIG. 4.

According to an embodiment of the present disclosure, the memory 130 is a flash memory type, hard disk type, multimedia card micro type, or card type memory (e.g. (e.g. SD or -Only Memory), and may include at least one type of storage medium among magnetic memory, magnetic disk, and optical disk. The computing device 100 may operate in connection with web storage that performs a storage function of the memory 130 on the Internet. The description of the memory described above is merely an example, and the present disclosure is not limited thereto.

The network unit 150 according to an embodiment of the present disclosure includes Public Switched Telephone Network (PSTN), x Digital Subscriber Line (xDSL), Rate Adaptive DSL (RADSL), Multi Rate DSL (MDSL), and VDSL ( A variety of wired communication systems can be used, such as Very High Speed DSL), Universal Asymmetric DSL (UADSL), High Bit Rate DSL (HDSL), and Local Area Network (LAN).

In addition, the network unit 150 presented in this specification includes Code Division Multi Access (CDMA), Time Division Multi Access (TDMA), Frequency Division Multi Access (FDMA), Orthogonal Frequency Division Multi Access (OFDMA), and SC-FDMA ( A variety of wireless communication systems can be used, such as Single Carrier-FDMA) and other systems.

In the present disclosure, the network unit 150 may use any type of wired or wireless communication system.

The techniques described herein can be used in the networks mentioned above, as well as other networks.

Figure 2 is a schematic diagram showing a network function according to an embodiment of the present disclosure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. A neural network can generally consist of a set of interconnected computational units, which can be referred to as nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node. Nodes (or neurons) that make up neural networks may be interconnected by one or more links.

Within a neural network, one or more nodes connected through a link may form a relative input node and output node relationship. The concepts of input node and output node are relative, and any node in an output node relationship with one node may be in an input node relationship with another node, and vice versa. As described above, input node to output node relationships can be created around links. One or more output nodes can be connected to one input node through a link, and vice versa.

In a relationship between an input node and an output node connected through one link, the value of the data of the output node may be determined based on the data input to the input node. Here, the link connecting the input node and the output node may have a weight. Weights may be variable and may be varied by the user or algorithm in order for the neural network to perform the desired function. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to the values input to the input nodes connected to the output node and the links corresponding to each input node. The output node value can be determined based on the weight.

As described above, in a neural network, one or more nodes are interconnected through one or more links to form an input node and output node relationship within the neural network. The characteristics of the neural network may be determined according to the number of nodes and links within the neural network, the correlation between the nodes and links, and the value of the weight assigned to each link. For example, if the same number of nodes and links exist and two neural networks with different weight values of the links exist, the two neural networks may be recognized as different from each other.

A neural network may consist of a set of one or more nodes. A subset of nodes that make up a neural network can form a layer. Some of the nodes constituting the neural network may form one layer based on the distances from the first input node. For example, a set of nodes with a distance n from the initial input node may constitute n layers. The distance from the initial input node can be defined by the minimum number of links that must be passed to reach the node from the initial input node. However, this definition of a layer is arbitrary for explanation purposes, and the order of a layer within a neural network may be defined in a different way than described above. For example, a layer of nodes may be defined by distance from the final output node.

The initial input node may refer to one or more nodes in the neural network through which data is directly input without going through links in relationships with other nodes. Alternatively, in the relationship between nodes based on links within a neural network, it may refer to nodes that do not have other input nodes connected by links. Similarly, the final output node may refer to one or more nodes that do not have an output node in their relationship with other nodes among the nodes in the neural network. Additionally, hidden nodes may refer to nodes constituting a neural network other than the first input node and the last output node.

The neural network according to an embodiment of the present disclosure is a neural network in which the number of nodes in the input layer may be the same as the number of nodes in the output layer, and the number of nodes decreases and then increases again as it progresses from the input layer to the hidden layer. You can. In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer may be less than the number of nodes in the output layer, and the number of nodes decreases as it progresses from the input layer to the hidden layer. there is. In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer may be greater than the number of nodes in the output layer, and the number of nodes increases as it progresses from the input layer to the hidden layer. You can. A neural network according to another embodiment of the present disclosure may be a neural network that is a combination of the above-described neural networks.

A deep neural network (DNN) may refer to a neural network that includes multiple hidden layers in addition to the input layer and output layer. Deep neural networks allow you to identify latent structures in data. In other words, it is possible to identify the potential structure of a photo, text, video, voice, or music (e.g., what object is in the photo, what the content and emotion of the text are, what the content and emotion of the voice are, etc.) . Deep neural networks include convolutional neural network (CNN), recurrent neural network (RNN), auto encoder, Generative Adversarial Networks (GAN), and Long Short-Term Memory. ; LSTM, Transformer, restricted Boltzmann machine (RBM), deep belief network (DBN), Q network, U network, Siamese network, Generative Adversarial Network (GAN) It may include etc. The description of the deep neural network described above is only an example and the present disclosure is not limited thereto.

In one embodiment of the present disclosure, the network function may include an autoencoder. An autoencoder may be a type of artificial neural network to output output data similar to input data. The autoencoder may include at least one hidden layer, and an odd number of hidden layers may be placed between input and output layers. The number of nodes in each layer may be reduced from the number of nodes in the input layer to an intermediate layer called the bottleneck layer (encoding), and then expanded symmetrically and reduced from the bottleneck layer to the output layer (symmetrical to the input layer). Autoencoders can perform nonlinear dimensionality reduction. The number of input layers and output layers can be corresponded to the dimension after preprocessing of the input data. In an auto-encoder structure, the number of nodes in the hidden layer included in the encoder may have a structure that decreases as the distance from the input layer increases. If the number of nodes in the bottleneck layer (the layer with the fewest nodes located between the encoder and decoder) is too small, not enough information may be conveyed, so if it is higher than a certain number (e.g., more than half of the input layers, etc.) ) may be maintained.

A neural network may be trained in at least one of supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning. Learning of a neural network may be a process of applying knowledge for the neural network to perform a specific operation to the neural network.

Neural networks can be trained to minimize output errors. In neural network learning, learning data is repeatedly input into the neural network, the output of the neural network and the error of the target for the learning data are calculated, and the error of the neural network is transferred from the output layer of the neural network to the input layer in the direction of reducing the error. This is the process of updating the weight of each node in the neural network through backpropagation. In the case of teacher learning, learning data in which the correct answer is labeled in each learning data is used (i.e., labeled learning data), and in the case of non-teacher learning, the correct answer may not be labeled in each learning data. That is, for example, in the case of teacher learning regarding data classification, the learning data may be data in which each learning data is labeled with a category. Labeled training data is input to the neural network, and the error can be calculated by comparing the output (category) of the neural network with the label of the training data. As another example, in the case of non-teachable learning for data classification, the error can be calculated by comparing the input training data with the neural network output. The calculated error is backpropagated in the reverse direction (i.e., from the output layer to the input layer) in the neural network, and the connection weight of each node in each layer of the neural network can be updated according to backpropagation. The amount of change in the connection weight of each updated node may be determined according to the learning rate. The neural network's calculation of input data and backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently depending on the number of repetitions of the learning cycle of the neural network. For example, in the early stages of neural network training, a high learning rate can be used to increase efficiency by allowing the neural network to quickly achieve a certain level of performance, and in the later stages of training, a low learning rate can be used to increase accuracy.

In the learning of neural networks, the training data can generally be a subset of real data (i.e., the data to be processed using the learned neural network), and thus the error for the training data is reduced, but the error for the real data is reduced. There may be an incremental learning cycle. Overfitting is a phenomenon in which errors in actual data increase due to excessive learning on training data. For example, a phenomenon in which a neural network that learned a cat by showing a yellow cat fails to recognize that it is a cat when it sees a non-yellow cat may be a type of overfitting. Overfitting can cause errors in machine learning algorithms to increase. To prevent such overfitting, various optimization methods can be used. To prevent overfitting, methods such as increasing the learning data, regularization, dropout to disable some of the network nodes during the learning process, and use of a batch normalization layer can be applied. You can.

Figure 3 is a flowchart showing a process of performing prediction based on a graph neural network model according to an embodiment of the present disclosure.

As described above in the background, in a specific node classification problem, there is a need to reflect the frequency of occurrence of each class into which the node is classified or the topological distance information between nodes associated with the classes.

Taking the structure of the compound shown in Figure 5 as an example, in order for a compound to bind to a kinase, the compound must generally have a detailed structure capable of hydrogen bonding with the hinge region of the kinase. The part in the hinge of the kinase that forms a hydrogen bond with the compound is based on the gate-keeper residue (GK), the backbone carbonyl oxygen (hydrogen bond acceptor) of the next residue (GK+1) of the gate-keeper residue, and the gate-keeper residue. The backbone carbonyl oxygen (hydrogen bond acceptor) and amide nitrogen (hydrogen bond donor) of the third residue (GK+3) from Accordingly, since the compound forms hydrogen bonds with the kinase hinge region at up to three positions, the frequency of occurrence of the class must be limited in the problem of predicting which atoms in a compound hydrogen bond with the hinge region of the kinase. Additionally, the topological distance between nodes classified into each class must be considered.

According to FIG. 3, the process of performing prediction based on the graph neural network model according to the present disclosure is to create a label based on a label value matrix containing label values of a plurality of nodes and a matrix containing distance information between a plurality of nodes. A step of calculating a class information matrix (S310), a step of calculating a predicted class information matrix based on a prediction value matrix including prediction values of a plurality of nodes and a matrix including distance information between a plurality of nodes (S320) , a step of calculating the norm between the label class information matrix and the predicted class information matrix (S330), a step of learning a graph neural network model in the direction where the calculated norm value is minimized (S340), and the learned It includes a step (S350) of performing a prediction based on the input information using a graph neural network model.

In step S310, a label class information matrix is calculated based on a label value matrix including label values of a plurality of nodes and a matrix including distance information between the plurality of nodes.

The label value matrix containing the label values of the nodes may be a matrix for each node and class information labeled on the nodes.

For example, in the case of a node classification problem of a compound that binds to the hinge region of a kinase, a label value matrix containing the label values of a plurality of nodes provides class information to each node of the compound, that is, the binding position of the kinase that binds to the compound. The information may be a labeled matrix. Specifically, when explaining using Figures 5 and 6, the atoms constituting the compound shown in Figure 5 can be classified into four classes according to the binding position of the kinase. That is, among the N (nitrogen atom) and O (oxygen atom) in the compound, from the amino acid residue corresponding to the gate keeper (GK) to the backbone carbonyl oxygen of the amino acid residue closest to the first (GK+1) The atom that hydrogen bonds to is class 1, and the atom that hydrogen bonds to the backbone amide nitrogen of the amino acid residue third (GK+3) closest to the amino acid residue corresponding to the gatekeeper among N and O in the compound is class 2, the compound. Among the N and O in the compound, the atom that hydrogen bonds to the backbone carbonyl oxygen of the third (GK+3) closest amino acid residue from the amino acid residue corresponding to the gatekeeper is class 3, and among the N and O in the compound, it corresponds to classes 1 to 3. Atoms that do not can be classified as class 4.

Candidate atoms that can bind to the hinge region of the kinase in the compound shown in FIG. 5 are atoms 1, 5, 8, 10, 15, 19, 25, 35, 37, and 39. When expressing whether or not it binds to the binding site of a kinase as a one-hot vector, the label value matrix has information about each atom as a row and information about whether or not it binds to the binding site of a kinase as a column. , it can be a matrix with a size of (number of nodes Figure 6 shows the label value matrix corresponding to the information.

A matrix containing distance information between a plurality of nodes may be a matrix containing topological distance information between nodes associated with a plurality of classes into which the nodes are classified. For example, in the example of the node classification problem of a compound, the matrix containing distance information may be a matrix with a size (number of nodes

, 복수의 노드들 사이의 거리 정보를 포함하는 행렬을

라고 할 때, 라벨 클래스 정보 행렬

은 [수학식 1]과 같이 연산될 수 있다.A label value matrix containing the label values of multiple nodes.

, a matrix containing distance information between a plurality of nodes

When we say that, the label class information matrix

can be calculated as in [Equation 1].

는 i번째 클래스의 출현 빈도를 의미하고, 비대각 원소(off-diagonal element)인

는 i번째 클래스의 노드들과 j번째 클래스의 노드들의 위상학적 거리의 합을 의미한다.At this time, the R matrix is a matrix with a size of (number of classes

means the frequency of occurrence of the ith class, and is an off-diagonal element.

means the sum of the topological distances of the nodes of the i-th class and the nodes of the j-th class.

For a specific example with reference to Figure 5, in Figure 5, there are three positions in the hinge region of the kinase that can form hydrogen bonds with the compound, and a maximum of three atoms are bonded to the hinge region of the kinase, so the Each node corresponding to an atom can be classified into three classes. At this time, the label matrix R can be calculated as in [Equation 2].

은 각 3개의 클래스의 출현 빈도를 의미하고, 비대각 원소인

은 각각 클래스1의 노드들과 클래스2의 노드들의 위상학적 거리의 합, 클래스1의 노드들과 클래스3의 노드들의 위상학적 거리의 합, 클래스2의 노드들과 클래스3의 노드들의 위상학적 거리의 합을 의미할 수 있다.At this time, the diagonal elements of the matrix are

means the frequency of occurrence of each of the three classes, and is an off-diagonal element

is the sum of the topological distances of the nodes of class 1 and the nodes of class 2, the sum of the topological distances of the nodes of class 1 and the nodes of class 3, and the topological distance of the nodes of class 2 and the nodes of class 3, respectively. It can mean the sum of .

, 복수의 노드들 사이의 거리 정보를 포함하는 행렬을

라고 할 때, 예측 클래스 정보 행렬

은 [수학식 3]과 같이 연산될 수 있다.In step S320, the processor 110 may calculate a prediction class information matrix based on a prediction value matrix including prediction values of a plurality of nodes and a matrix including distance information between the plurality of nodes. A prediction value matrix containing information about the probability that a plurality of nodes will be classified into each class, that is, prediction values,

, a matrix containing distance information between a plurality of nodes

When we say, the predicted class information matrix

can be calculated as in [Equation 3].

의 각 성분은 각 노드의 예측된 로짓(logit)값을 소프트맥스(softmax)활성함수로 변형시킨 값일 수 있으나, 본 개시는 이에 한정하지 아니한다.procession

Each component of may be a value obtained by transforming the predicted logit value of each node into a softmax activation function, but the present disclosure is not limited to this.

및 예측 클래스 정보 행렬

사이의 노름(norm)을 연산하여, 그 값을 그래프 인공 신경망의 손실 함수로 지정할 수 있다. 이 때 그래프 신경망의 손실 함수는 [수학식 4]와 같이 연산될 수 있다.In step S330, the processor 110 uses the label class information matrix

and predicted class information matrix

By calculating the norm between the values, the value can be specified as the loss function of the graph artificial neural network. At this time, the loss function of the graph neural network can be calculated as in [Equation 4].

In step S340, the processor 110 may train the graph neural network model in a direction that minimizes the loss function calculated in step S330. The loss function includes a label class information matrix and a prediction class information matrix, and each class information matrix includes information about the frequency of occurrence of the class and information about the topological distance between nodes of the class. Therefore, when learning a graph neural network model using [Equation 5] as the loss function, it is possible to train it by reflecting the constraints on the frequency of occurrence of the class and the constraints on the topological distance.

In another embodiment, instead of calculating the norm for the entire label class information matrix and the entire prediction class information matrix, each element of the matrix is separated and each norm is calculated, thereby creating a loss function tailored to various problems. can be designed. For example, design a loss function that considers the frequency of occurrence of each class by calculating the L2 norm using only the diagonal elements of each matrix, or calculate the L2 norm using only the off-diagonal elements of each matrix. By calculating , you can design a loss function that considers topological distance information between nodes associated with each class. However, the present disclosure is not limited to the above example, and various types of loss functions can be derived from the equation.

In step S350, the processor 110 may perform prediction based on the input information using the graph neural network model learned in step S340. Following the above example of a compound and a kinase, the processor 110 can receive information about the structure of the compound using a learned graph neural network model and predict which of the hinge regions of the kinase each node binds to.

In one embodiment of the present disclosure, the processor 110 designates the loss function calculated based on the label class information matrix and the prediction class information matrix as the first loss function as above, and generates a cross entropy (cross entropy) together with the first loss function. A graph neural network model can be trained by using a second loss function calculated based on entropy. Since the cross-entropy loss function is a loss function mainly used to improve the performance of graph neural networks, the performance of the final trained graph neural network can be higher by using the two types of loss functions together. At this time, the final loss function for training the model can be expressed as [Equation 5].

는 크로스 엔트로피 손실 함수,

는 본 개시에 따른 클래스들 각각의 발생 빈도 및 거리 정보가 반영된 손실 함수일 수 있다.At this time,

is the cross-entropy loss function,

may be a loss function that reflects the frequency of occurrence and distance information of each class according to the present disclosure.

In another embodiment, a loss function for training a graph neural network model may be determined by calculating a weighted sum of a first loss function and a second loss function. At this time, the final loss function for training the model can be expressed as [Equation 6].

는 제2 손실 함수 즉 크로스 엔트로피 손실 함수의 가중치,

는 크로스 엔트로피 손실 함수,

는 본 개시에 따른 클래스들 각각의 발생 빈도 및 거리 정보가 반영된 손실 함수의 가중치,

는 본 개시에 따른 클래스들 각각의 발생 빈도 및 클래스들과 연관된 거리 정보가 반영된 손실 함수일 수 있다.At this time,

is the weight of the second loss function, that is, the cross-entropy loss function,

is the cross-entropy loss function,

is the weight of the loss function reflecting the frequency of occurrence and distance information of each class according to the present disclosure,

May be a loss function that reflects the frequency of occurrence of each class and distance information associated with the classes according to the present disclosure.

As described above, in this disclosure, when a graph neural network model is trained using a loss function reflecting the frequency of occurrence of each class and the distance information associated with the classes together with a cross-entropy loss function, the graph neural network applies constraints according to the frequency of occurrence and distance. While it can be learned to reflect better, the contribution or contribution timing of each loss function can be adjusted by adjusting or scheduling the weight assigned to each loss function.

For example, in the problem of predicting the location of the atom that hydrogen bonds with a kinase among the atoms constituting a compound, the class asymmetry of the label data is very large. In other words, the data of a specific class is significantly more than the data of other classes. When such class asymmetry exists, training a graph neural network using only the cross-entropy function shows relatively low prediction values. On the other hand, when applying a loss function that reflects the frequency of occurrence and distance information of each class as in the present disclosure, the prediction accuracy of the graph neural network model can be improved.

In addition, since the loss function that reflects the frequency of occurrence and distance information of each class of the present disclosure can improve the performance of the graph neural network model in both cases where asymmetry is large and small, This can solve the so-called accuracy paradox phenomenon, in which model accuracy improves but prediction accuracy is sacrificed for classes with relatively small class asymmetry.

Ultimately, through this disclosure, the performance of the graph neural network model can be further improved to perform more accurate predictions.

Meanwhile, according to an embodiment of the present disclosure, a method for learning a graph neural network model is disclosed.

Figure 4 is a flowchart showing a method of learning a graph neural network model according to an embodiment of the present disclosure.

The learning method of the graph neural network model according to FIG. 4 includes calculating a first class information matrix related to the label values of a plurality of nodes (S410), and calculating a second class information matrix related to the predicted values of a plurality of nodes (S410). S420), calculating a loss function based on the first class information matrix and the second class information matrix (S430), and a method of training a graph neural network model based on the loss function.

In step S410, the processor 110 may calculate a first class information matrix related to label values of a plurality of nodes. Specifically, the processor 110 may calculate the first class information matrix based on a label value matrix including label values of a plurality of nodes and a distance matrix including distance information between a plurality of nodes. The method of calculating the first class information matrix may correspond to the method of calculating the label class information matrix of FIG. 3. A specific method of calculating the label class information matrix was described in detail with reference to FIG. 3.

In step S420, the processor 110 may calculate a second class information matrix related to the predicted values of the plurality of nodes. Specifically, the processor 110 may calculate a second class information matrix based on a prediction value matrix including prediction values of a plurality of nodes and a distance matrix. The method of calculating the second class information matrix may correspond to the method of calculating the predicted class information matrix of FIG. 3. A specific method of calculating the prediction class information matrix was described in detail with reference to FIG. 3.

Each of the first class information matrix and the second class information matrix calculated in steps S410 and S420 may include the frequency of occurrence of each of the plurality of classes and distance information associated with the plurality of classes.

In step S430, the processor 110 may calculate a loss function based on the first class information matrix and the second class information matrix. At this time, the loss function can be calculated using the norm values of the first class information matrix and the second class information matrix.

As described above, when a graph neural network model is trained using a loss function reflecting the frequency of occurrence of each class and distance information associated with the classes and a cross-entropy loss function through the learning method of the graph neural network model of the present disclosure, the learned graph Neural networks can better reflect constraints based on frequency of occurrence and distance. Therefore, the performance of a graph neural network learned by the learning method of the graph neural network model of the present disclosure can be improved.

Meanwhile, according to an embodiment of the present disclosure, a computer-readable medium storing a data structure is disclosed.

Data structure can refer to the organization, management, and storage of data to enable efficient access and modification of data. Data structure can refer to the organization of data to solve a specific problem (e.g., retrieving data, storing data, or modifying data in the shortest possible time). A data structure may be defined as a physical or logical relationship between data elements designed to support a specific data processing function. Logical relationships between data elements may include connection relationships between user-defined data elements. Physical relationships between data elements may include actual relationships between data elements that are physically stored in a computer-readable storage medium (e.g., a persistent storage device). A data structure may specifically include a set of data, relationships between data, and functions or instructions applicable to the data. Effectively designed data structures allow computing devices to perform computations while minimizing the use of the computing device's resources. Specifically, computing devices can increase the efficiency of operations, reading, insertion, deletion, comparison, exchange, and search through effectively designed data structures.

Data structures can be divided into linear data structures and non-linear data structures depending on the type of data structure. A linear data structure may be a structure in which only one piece of data is connected to another piece of data. Linear data structures may include List, Stack, Queue, and Deque. A list can refer to a set of data that has an internal order. The list may include a linked list. A linked list may be a data structure in which data is connected in such a way that each data is connected in a single line with a pointer. In a linked list, a pointer may contain connection information to the next or previous data. Depending on its form, a linked list can be expressed as a singly linked list, a doubly linked list, or a circularly linked list. A stack may be a data listing structure that allows limited access to data. A stack can be a linear data structure in which data can be processed (for example, inserted or deleted) at only one end of the data structure. Data stored in the stack may have a data structure (LIFO-Last in First Out) where the later it enters, the sooner it comes out. A queue is a data listing structure that allows limited access to data. Unlike the stack, it can be a data structure (FIFO-First in First Out) where data stored later is released later. A deck can be a data structure that can process data at both ends of the data structure.

A non-linear data structure may be a structure in which multiple pieces of data are connected behind one piece of data. Nonlinear data structures may include graph data structures. A graph data structure can be defined by vertices and edges, and an edge can include a line connecting two different vertices. Graph data structure may include a tree data structure. A tree data structure may be a data structure in which there is only one path connecting two different vertices among a plurality of vertices included in the tree. In other words, it may be a data structure that does not form a loop in the graph data structure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. Below, it is described in a unified manner as a neural network. Data structures may include neural networks. And the data structure including the neural network may be stored in a computer-readable medium. Data structures including neural networks also include data preprocessed for processing by a neural network, data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, neural network It may include a loss function for learning. A data structure containing a neural network may include any of the components disclosed above. In other words, the data structure including the neural network includes data preprocessed for processing by the neural network, data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, neural network It may be configured to include all or any combination of the loss function for learning. In addition to the configurations described above, a data structure containing a neural network may include any other information that determines the characteristics of the neural network. Additionally, the data structure may include all types of data used or generated in the computational process of a neural network and is not limited to the above. Computer-readable media may include computer-readable recording media and/or computer-readable transmission media. A neural network can generally consist of a set of interconnected computational units, which can be referred to as nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node.

The data structure may include data input to the neural network. A data structure containing data input to a neural network may be stored in a computer-readable medium. Data input to the neural network may include learning data input during the neural network learning process and/or input data input to the neural network on which training has been completed. Data input to the neural network may include data that has undergone pre-processing and/or data subject to pre-processing. Preprocessing may include a data processing process to input data into a neural network. Therefore, the data structure may include data subject to preprocessing and data generated by preprocessing. The above-described data structure is only an example and the present disclosure is not limited thereto.

The data structure may include the weights of the neural network. (In this specification, weights and parameters may be used with the same meaning.) And the data structure including the weights of the neural network may be stored in a computer-readable medium. A neural network may include multiple weights. Weights may be variable and may be varied by the user or algorithm in order for the neural network to perform the desired function. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to the values input to the input nodes connected to the output node and the links corresponding to each input node. Based on the weight, the data value output from the output node can be determined. The above-described data structure is only an example and the present disclosure is not limited thereto.

As an example and not a limitation, the weights may include weights that are changed during the neural network learning process and/or weights for which neural network learning has been completed. Weights that change during the neural network learning process may include weights that change at the start of the learning cycle and/or weights that change during the learning cycle. Weights for which neural network training has been completed may include weights for which a learning cycle has been completed. Therefore, the data structure including the weights of the neural network may include weights that are changed during the neural network learning process and/or the data structure including the weights for which neural network learning has been completed. Therefore, the above-mentioned weights and/or combinations of each weight are included in the data structure including the weights of the neural network. The above-described data structure is only an example and the present disclosure is not limited thereto.

The data structure including the weights of the neural network may be stored in a computer-readable storage medium (e.g., memory, hard disk) after going through a serialization process. Serialization can be the process of converting a data structure into a form that can be stored on the same or a different computing device and later reconstituted and used. Computing devices can transmit and receive data over a network by serializing data structures. Data structures containing the weights of a serialized neural network can be reconstructed on the same computing device or on a different computing device through deserialization. The data structure including the weights of the neural network is not limited to serialization. Furthermore, the data structure including the weights of the neural network is a data structure to increase computational efficiency while minimizing the use of computing device resources (e.g., in non-linear data structures, B-Tree, Trie, m-way search tree, AVL tree, Red-Black Tree) may be included. The foregoing is merely an example and the present disclosure is not limited thereto.

The data structure may include hyper-parameters of a neural network. And the data structure including the hyperparameters of the neural network can be stored in a computer-readable medium. A hyperparameter may be a variable that can be changed by the user. Hyperparameters include, for example, learning rate, cost function, number of learning cycle repetitions, weight initialization (e.g., setting the range of weight values subject to weight initialization), Hidden Unit. It may include a number (e.g., number of hidden layers, number of nodes in hidden layers). The above-described data structure is only an example and the present disclosure is not limited thereto.

7 is a brief, general schematic diagram of an example computing environment in which embodiments of the present disclosure may be implemented.

Although the present disclosure has generally been described above as being capable of being implemented by a computing device, those skilled in the art will understand that the present disclosure can be implemented in combination with computer-executable instructions and/or other program modules that can be executed on one or more computers and/or in hardware and software. It will be well known that it can be implemented as a combination.

Typically, program modules include routines, programs, components, data structures, etc. that perform specific tasks or implement specific abstract data types. Additionally, those skilled in the art will understand that the methods of the present disclosure are applicable to single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, etc. It will be appreciated that each of these may be implemented in other computer system configurations, including those capable of operating in conjunction with one or more associated devices.

The described embodiments of the disclosure can also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Computers typically include a variety of computer-readable media. Computer-readable media can be any medium that can be accessed by a computer, and such computer-readable media includes volatile and non-volatile media, transitory and non-transitory media, removable and non-transitory media. Includes removable media. By way of example, and not limitation, computer-readable media may include computer-readable storage media and computer-readable transmission media. Computer-readable storage media refers to volatile and non-volatile media, transient and non-transitory media, removable and non-removable, implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Includes media. Computer-readable storage media may include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage. This includes, but is not limited to, a device, or any other medium that can be accessed by a computer and used to store desired information.

A computer-readable transmission medium typically implements computer-readable instructions, data structures, program modules, or other data on a modulated data signal, such as a carrier wave or other transport mechanism. Includes all information delivery media. The term modulated data signal refers to a signal in which one or more of the characteristics of the signal have been set or changed to encode information within the signal. By way of example, and not limitation, computer-readable transmission media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above are also intended to be included within the scope of computer-readable transmission media.

An example environment 1100 is shown that implements various aspects of the present disclosure, including a computer 1102, which includes a processing unit 1104, a system memory 1106, and a system bus 1108. do. System bus 1108 couples system components, including but not limited to system memory 1106, to processing unit 1104. Processing unit 1104 may be any of a variety of commercially available processors. Dual processors and other multiprocessor architectures may also be used as processing unit 1104.

System bus 1108 may be any of several types of bus structures that may further be interconnected to a memory bus, peripheral bus, and local bus using any of a variety of commercial bus architectures. System memory 1106 includes read only memory (ROM) 1110 and random access memory (RAM) 1112. The basic input/output system (BIOS) is stored in non-volatile memory 1110, such as ROM, EPROM, and EEPROM, and is a basic input/output system that helps transfer information between components within the computer 1102, such as during startup. Contains routines. RAM 1112 may also include high-speed RAM, such as static RAM, for caching data.

Computer 1102 may also include an internal hard disk drive (HDD) 1114 (e.g., EIDE, SATA)—the internal hard disk drive 1114 may also be configured for external use within a suitable chassis (not shown). Yes - a magnetic floppy disk drive (FDD) 1116 (e.g., for reading from or writing to a removable diskette 1118), and an optical disk drive 1120 (e.g., a CD-ROM for reading the disk 1122 or reading from or writing to other high-capacity optical media such as DVDs). Hard disk drive 1114, magnetic disk drive 1116, and optical disk drive 1120 are connected to system bus 1108 by hard disk drive interface 1124, magnetic disk drive interface 1126, and optical drive interface 1128, respectively. ) can be connected to. The interface 1124 for implementing an external drive includes at least one or both of Universal Serial Bus (USB) and IEEE 1394 interface technologies.

These drives and their associated computer-readable media provide non-volatile storage of data, data structures, computer-executable instructions, and the like. For computer 1102, drive and media correspond to storing any data in a suitable digital format. Although the description of computer-readable media above refers to removable optical media such as HDDs, removable magnetic disks, and CDs or DVDs, those skilled in the art will also recognize removable optical media such as zip drives, magnetic cassettes, flash memory cards, cartridges, etc. It will be appreciated that other types of computer-readable media, such as the like, may also be used in the example operating environment and that any such media may contain computer-executable instructions for performing the methods of the present disclosure.

A number of program modules may be stored in the drive and RAM 1112, including an operating system 1130, one or more application programs 1132, other program modules 1134, and program data 1136. All or portions of the operating system, applications, modules and/or data may also be cached in RAM 1112. It will be appreciated that the present disclosure may be implemented on various commercially available operating systems or combinations of operating systems.

A user may enter commands and information into computer 1102 through one or more wired/wireless input devices, such as a keyboard 1138 and a pointing device such as mouse 1140. Other input devices (not shown) may include microphones, IR remote controls, joysticks, game pads, stylus pens, touch screens, etc. These and other input devices are connected to the processing unit 1104 through an input device interface 1142, which is often connected to the system bus 1108, but may also include a parallel port, an IEEE 1394 serial port, a game port, a USB port, an IR interface, It can be connected by other interfaces, etc.

A monitor 1144 or other type of display device is also connected to system bus 1108 through an interface, such as a video adapter 1146. In addition to monitor 1144, computers typically include other peripheral output devices (not shown) such as speakers, printers, etc.

Computer 1102 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1148, via wired and/or wireless communications. Remote computer(s) 1148 may be a workstation, computing device computer, router, personal computer, portable computer, microprocessor-based entertainment device, peer device, or other conventional network node, and is generally connected to computer 1102. For simplicity, only memory storage device 1150 is shown, although it includes many or all of the components described. The logical connections depicted include wired/wireless connections to a local area network (LAN) 1152 and/or a larger network, such as a wide area network (WAN) 1154. These LAN and WAN networking environments are common in offices and companies and facilitate enterprise-wide computer networks, such as intranets, all of which can be connected to a worldwide computer network, such as the Internet.

When used in a LAN networking environment, computer 1102 is connected to local network 1152 through wired and/or wireless communication network interfaces or adapters 1156. Adapter 1156 may facilitate wired or wireless communication to LAN 1152, which also includes a wireless access point installed thereon for communicating with wireless adapter 1156. When used in a WAN networking environment, the computer 1102 may include a modem 1158 or be connected to a communicating computing device on the WAN 1154 or to establish communications over the WAN 1154, such as over the Internet. Have other means. Modem 1158, which may be internal or external and a wired or wireless device, is coupled to system bus 1108 via serial port interface 1142. In a networked environment, program modules described for computer 1102, or portions thereof, may be stored in remote memory/storage device 1150. It will be appreciated that the network connections shown are exemplary and that other means of establishing a communications link between computers may be used.

Computer 1102 may be associated with any wireless device or object deployed and operating in wireless communications, such as a printer, scanner, desktop and/or portable computer, portable data assistant (PDA), communications satellite, wirelessly detectable tag. Performs actions to communicate with any device or location and telephone. This includes at least Wi-Fi and Bluetooth wireless technologies. Accordingly, communication may be a predefined structure as in a conventional network or may simply be ad hoc communication between at least two devices.

Wi-Fi (Wireless Fidelity) allows connection to the Internet, etc. without wires. Wi-Fi is a wireless technology, like cell phones, that allows these devices, such as computers, to send and receive data indoors and outdoors, anywhere within the coverage area of a cell tower. Wi-Fi networks use wireless technology called IEEE 802.11 (a, b, g, etc.) to provide secure, reliable, and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, to the Internet, and to wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in the unlicensed 2.4 and 5 GHz wireless bands, for example, at data rates of 11 Mbps (802.11a) or 54 Mbps (802.11b), or in products that include both bands (dual band). .

Those skilled in the art will understand that information and signals may be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols and chips that may be referenced in the above description include voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields. It can be expressed by particles or particles, or any combination thereof.

Those skilled in the art will understand that the various illustrative logical blocks, modules, processors, means, circuits and algorithm steps described in connection with the embodiments disclosed herein may be used in electronic hardware, (for convenience) It will be understood that it may be implemented by various forms of program or design code (referred to herein as software) or a combination of both. To clearly illustrate this interoperability of hardware and software, various illustrative components, blocks, modules, circuits and steps have been described above generally with respect to their functionality. Whether this functionality is implemented as hardware or software depends on the specific application and design constraints imposed on the overall system. A person skilled in the art of this disclosure may implement the described functionality in various ways for each specific application, but such implementation decisions should not be construed as departing from the scope of this disclosure.

The various embodiments presented herein may be implemented as a method, apparatus, or article of manufacture using standard programming and/or engineering techniques. The term article of manufacture includes a computer program, carrier, or media accessible from any computer-readable storage device. For example, computer-readable storage media include magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash. Includes, but is not limited to, memory devices (e.g., EEPROM, cards, sticks, key drives, etc.). Additionally, various storage media presented herein include one or more devices and/or other machine-readable media for storing information.

It is to be understood that the specific order or hierarchy of steps in the processes presented is an example of illustrative approaches. It is to be understood that the specific order or hierarchy of steps in processes may be rearranged within the scope of the present disclosure, based on design priorities. The appended method claims present elements of the various steps in a sample order but are not meant to be limited to the particular order or hierarchy presented.

The description of the presented embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope of the disclosure. Thus, the present disclosure is not limited to the embodiments presented herein but is to be interpreted in the broadest scope consistent with the principles and novel features presented herein.

Claims (13)

A method performed by a computing device to make a prediction based on a graph neural network model, comprising:
Inputting information into the graph neural network model; and
performing prediction based on the input information using the graph neural network model;
Including,
The graph neural network model is,
The frequency of occurrence of each of a plurality of classes related to the classification of the node, and
Corresponding to a model learned using a loss function calculated based on distance information associated with the plurality of classes,
method.
According to claim 1,
The predictions include predictions related to the binding potential of the compound with the kinase,
The node corresponds to an atom included in the compound,
method.
According to claim 2,
The plurality of classes are associated with a plurality of binding sites of the kinase,
method.
According to claim 1,
Distance information associated with the plurality of classes is,
Topological distance information between nodes associated with the plurality of classes;
Including,
method.
According to claim 1,
The loss function is,
A first loss function calculated based on the norm between the class information matrix related to the label value of the node included in the learning data and the class information matrix related to the predicted value of the node
Including,
method.
According to claim 5,
The class information matrix related to the label value of the node is,
Calculated based on a label value matrix including label values of a plurality of nodes and a distance matrix including distance information between the plurality of nodes,
The class information matrix related to the predicted value of the node is:
Calculated based on a prediction value matrix including prediction values of the plurality of nodes and a distance matrix including distance information between the plurality of nodes,
method.
According to claim 5,
Each class information matrix is,
A diagonal element containing information about the frequency of occurrence of each class; and
An off-diagonal element containing topological distance information between nodes associated with different classes;
Including,
method.
According to claim 5,
The loss function is,
a second loss function calculated based on the cross entropy of the learning data;
Containing more,
method.
According to claim 8,
The loss function is,
Determined by a weighted sum operation of the first loss function and the second loss function,
method.
A computer program stored on a computer-readable storage medium containing instructions that cause a computing device to perform operations, the operations comprising:
An operation of inputting information into a graph neural network model; and
An operation of performing prediction based on the input information using the graph neural network model;
Including,
The graph neural network model is,
The frequency of occurrence of each of a plurality of classes related to the classification of the node, and
Distance information associated with the plurality of classes
Corresponding to the model learned using the loss function calculated based on
method.
As a computing device,
at least one processor; and
Memory;
Including,
The at least one processor,
Enter information into the graph neural network model,
Using the graph neural network model, prediction is performed based on the input information,
The graph neural network model is,
The frequency of occurrence of each of a plurality of classes related to the classification of the node, and
Distance information associated with the plurality of classes
Corresponding to the model learned using the loss function calculated based on
Computing device.
As a learning method for a graph neural network model,
calculating a first class information matrix related to label values of a plurality of nodes;
calculating a second class information matrix related to predicted values of the plurality of nodes;
calculating a loss function based on the first class information matrix and the second class information matrix; and
training the graph neural network model based on the loss function;
Including,
Each of the first class information matrix and the second class information matrix is:
The frequency of occurrence of each of the plurality of classes, and
Distance information associated with the plurality of classes
Including,
Learning methods for graph neural network models.
According to claim 12,
The step of calculating the first class information matrix is,
Comprising a step of calculating a first class information matrix based on a label value matrix including label values of the plurality of nodes, and a distance matrix including distance information between the plurality of nodes,
The step of calculating the second class information matrix is,
Comprising a prediction value matrix including prediction values of the plurality of nodes, and calculating a second class information matrix based on the distance matrix,
Learning methods for graph neural network models.

Images