Target protein structure-based candidate material discovery system for predicting binding affinity and operation method thereof

Abstract

This operation method of the target protein structure-based candidate material discovery system for predicting binding affinity includes the steps of: displaying a first screen provided with a plurality of ligands to be docked to a target protein; calculating predicted bonding energy while sequentially performing molecular docking on the plurality of ligands according to a setting inputted by a user through a first screen; displaying, on a second screen, a first list consisting of rows including a user interface control for the selection of a user, names of a plurality of ligands, and values of the calculated bonding energy; predicting binding affinity corresponding to a protein ligand binding pose used for calculating bonding energy for rows selected by a user by using a user interface control in a first list of a second screen; and displaying, on a third screen, a second list consisting of rows including names of the plurality of ligands, values of binding energies, and values of binding affinities.

Description

Target protein structure-based candidate substance discovery system for predicting binding affinity and its operating method

[0001] The disclosure relates to a protein structure-based candidate substance discovery system and its operating method, and more specifically, to a target protein structure-based candidate substance discovery system for predicting binding affinity and its operating method.

[0002] Proteins have limited degrees of freedom and a specific three-dimensional structure based on their amino acid sequence. A protein's function is determined by its specific three-dimensional structure. Therefore, once a target protein is identified, it is possible to target that specific structure and search for candidate substances that can bind to the protein's active site and modulate its function. Structure-based design of materials shortens the development process and allows for the development of substances that act only on the target protein, minimizing undesirable effects, or side effects.

[0003] As such, target protein structure-based screening of candidate compounds is generally used primarily for the discovery of new drug candidates. This is because, due to the nature of new drug development, it is a representative field that requires reducing the time and cost required for the discovery of initial lead compounds or candidates, and there must be no side effects. The discovery of new drug candidates is an early stage in new drug development, and protein structure-based in silico screening methods are attracting attention. Protein structure-based in silico screening is a method for identifying potential new drug candidates from a compound database based on the three-dimensional structure of a protein to which the new drug candidate can bind. In silico refers to computer programming in computer simulation or virtual experiments, and protein structure-based in silico screening can predict the binding affinity of a compound by simulating the interaction between a protein and a compound. Protein structure-based in silico screening can be useful for finding compounds that precisely fit the active site of a protein, as well as compounds in novel chemical space, compared to ligand-based screening, which relies on known ligands and is limited in finding entirely novel chemical structures because it searches for compounds similar to already known ligands.

[0004] In the discovery of protein structure-based candidate compounds, where these principles apply, there is a growing need to integrate big data analysis and artificial intelligence technologies with libraries of billions of compounds, utilizing a variety of analytical tools in parallel. To meet this demand, research labs at pharmaceutical companies and universities have been installing and operating various software. For example, the European Molecular Biology Laboratory (EMBL) has developed and offers downloadable tools for bioinformatics, or provides simple web applications.

[0005] Even when protein structure-based candidate compounds are discovered, it's difficult to immediately begin development for commercialization. For example, once a new drug candidate is discovered, it undergoes various cell and animal experiments to verify and optimize it, ultimately leading to clinical trials for human application. However, even if a new drug candidate is identified, users with only a basic understanding of biology and no computer skills often find it difficult to directly apply the discovery results to clinical trials. This is because the discovery results provided by new drug candidate software alone typically do not provide experimental concentration values. In such cases, even after the new drug candidate is discovered, additional experiments are necessary to determine the experimental concentration values. Typically, these additional experiments involve performing numerous serial dilutions of the candidate drug, analyzing the results to determine the therapeutic range deemed effective, which is time-consuming and costly.

[0006] The task to be solved is to provide a protein structure-based candidate discovery system that can discover candidate substances based on protein structure and predict binding affinity between target proteins and ligands and provide it to users through a cloud platform.

[0007] Another challenge we are trying to address is to provide a new drug candidate discovery system that can predict the binding affinity between a target protein and a ligand and provide it to users through a cloud platform, along with the discovery of new drug candidates.

[0008] According to one embodiment, a method for operating a protein structure-based candidate substance discovery system is provided, which is implemented as a cloud platform that provides users with functions or services necessary for protein structure-based candidate substance discovery in the form of a web service, the method comprising: displaying a first screen for receiving a plurality of ligands to be docked to a target protein; calculating an expected binding energy while sequentially performing molecular docking on the plurality of ligands according to settings input by a user through the first screen; displaying a first list, which is composed of rows including a user interface control for a user's selection, names of the plurality of ligands, and values ​​of the calculated binding energies, on a second screen; predicting a binding affinity corresponding to a protein-ligand binding pose used for the binding energy calculation, for rows selected by the user using the user interface control from the first list of the second screen; and a step of displaying a second list consisting of rows including the names of the plurality of ligands, the values ​​of the binding energy, and the values ​​of the binding affinity on a third screen.

[0009] In some embodiments, the value of the binding energy may be expressed as a value in units representing the amount of energy, and the value of the binding affinity may be expressed as a value in units of molar concentration.

[0010] In some embodiments, the value of the binding energy may be expressed in units of Kcal / mol, and the value of the binding affinity may be expressed in units of fM, pM, nM, μM, mM, or M.

[0011] In some embodiments, the step of displaying the second list on the third screen may include the step of displaying the rows forming the second list on the third screen in descending or ascending order based on the value of the binding affinity.

[0012] In some embodiments, the binding affinity may be predicted by an artificial intelligence model trained using experimental measurements including at least one of a dissociation constant Kd, an inhibition constant Ki, and a half maximal inhibitory concentration IC50 as training data.

[0013] In some embodiments, the artificial intelligence model may be trained using the experimental measurement values ​​and the binding structure data of the target protein and the ligand as the training data.

[0014] In some embodiments, the artificial intelligence model may include a convolution layer comprising filters encoding patterns to be identified to predict binding affinity between the target protein and the ligand, and a dense layer integrating features extracted through the convolution layer.

[0015] In some embodiments, the convolution layer may include three convolution layers having 64, 128, and 256 filters, respectively.

[0016] In some embodiments, the dense layer may include three dense layers having 1000, 500, and 200 neurons, respectively.

[0017] In some embodiments, the artificial intelligence model may include a CNN (Convolutional Neural Network) model or a ResNet 3D (Residual Network 3D) model.

[0018] According to one embodiment, a protein structure-based candidate substance discovery system is implemented as a cloud platform that provides users with functions or services necessary for protein structure-based candidate substance discovery in the form of a web service, and may include: a project management module that creates a project in which tasks for performing discovery of protein structure-based candidate substances can be added; a simulation management module that creates a simulation desired by the user on the created project; a simulation setting module that sets a simulation flow for the simulation input by the user using a simulation setting area including a task module selection area, the simulation setting module including a first object that is changed into a first node for calculating an expected binding energy according to docking of a target protein and a ligand when dragged and dropped onto a canvas area, and a second object that is changed into a second node for predicting a binding affinity corresponding to a protein-ligand binding pose used for calculating the binding energy calculated at the first node; and a simulation flow management module that manages information on nodes that can precede or follow in the simulation flow settings.

[0019] In some embodiments, when the run button displayed on the first node is clicked, a first screen may be displayed that provides a plurality of ligands to be docked to the target protein.

[0020] In some embodiments, the first node may sequentially perform molecular docking on the plurality of ligands according to settings input by the user through the first screen and calculate the expected binding energy.

[0021] In some embodiments, after the calculation of the binding energy is completed, a first list comprising rows containing user interface controls for user selection, names of the plurality of ligands, and values ​​of the calculated binding energy may be displayed on a second screen.

[0022] In some embodiments, the second node can predict binding affinities corresponding to protein-ligand binding poses used in the binding energy calculation for rows selected by the user using the user interface controls in the first list of the second screen.

[0023] In some embodiments, after the prediction of the binding affinity is completed, a second list consisting of rows including the names of the plurality of ligands, the values ​​of the binding energies, and the values ​​of the binding affinities may be displayed on a third screen.

[0024] In some embodiments, the display of the second list on the third screen may include displaying the rows forming the second list on the third screen in descending or ascending order based on the value of the binding affinity.

[0025] In some embodiments, the value of the binding energy may be expressed in units of Kcal / mol, and the value of the binding affinity may be expressed in units of fM, pM, nM, μM, mM, or M.

[0026] In some embodiments, the binding affinity may be predicted by an artificial intelligence model trained using experimental measurements including at least one of a dissociation constant Kd, an inhibition constant Ki, and an inhibition median IC50 as training data.

[0027] In some embodiments, the simulation flow management module may manage information about nodes that can precede or follow in the simulation flow setup through metadata.

[0028] According to the embodiments, the present invention provides functions and a user interface optimized for the discovery of candidate substances based on protein structures, and improves the difficulties in data sharing and consistent data management due to the fact that detailed tasks related to the discovery of candidate substances based on protein structures were provided as separate tools with high incompatibility with each other, and the simulation flow can be easily managed by creating, changing, and deleting nodes.

[0029] In conjunction with this, by predicting and providing the binding affinity between the target protein and the ligand to the user, it helps the user intuitively understand the results and allows the user to easily calculate the molecular weight of the candidate substance and the volume of the medium derived without additional experiments, allowing direct application to clinical trials such as cell experiments. Therefore, the time and cost consumed in determining the experimental concentration for the required candidate substance can be significantly reduced. In particular, since these experimental concentrations are provided through a cloud platform, users can easily access the experimental concentration values ​​from anywhere with an internet connection, and collaboration with other users is also convenient.

[0030] Figure 1 is a block diagram illustrating a protein structure-based candidate substance discovery system according to one embodiment.

[0031] FIGS. 2 to 6 are diagrams showing exemplary screens of a protein structure-based candidate substance discovery system according to one embodiment.

[0032] Figure 7 is a diagram showing an example of a simulation setting of a protein structure-based candidate material discovery system according to one embodiment.

[0033] FIGS. 8 to 10 are drawings illustrating the operation of a protein structure-based candidate material discovery system according to one embodiment.

[0034] FIG. 11 is a diagram illustrating an operation method of a protein structure-based candidate substance discovery system according to one embodiment.

[0035] Fig. 12 is a diagram showing an example of an implementation of an artificial intelligence model of a protein structure-based candidate substance discovery system according to one embodiment.

[0036] FIG. 13 is a block diagram illustrating a computing device according to one embodiment.

[0037] Below, with reference to the attached drawings, embodiments of the present invention are described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In addition, in the drawings, parts irrelevant to the description are omitted for clarity of description, and similar parts are designated with similar reference numerals throughout the specification.

[0038] Throughout the specification and claims, whenever a part is said to "include" a certain component, this does not mean that it excludes other components, but rather that it may include other components, unless otherwise stated.

[0039] Additionally, terms such as “part,” “unit,” and “module” described in the specification may mean a unit capable of processing at least one function or operation described in the specification, which may be implemented by hardware, software, or a combination of hardware and software.

[0040] In this specification, the term "candidate substance" or "protein structure-based candidate substance" means any substance derived using the target protein structure-based candidate substance discovery system provided in this specification, without limitation to the field, and preferably includes the field of pharmaceutical development, food development, and livestock material discovery, and more preferably means a candidate substance in the field of pharmaceutical development, but is not limited thereto.

[0041] Figure 1 is a block diagram illustrating a protein structure-based candidate substance discovery system according to one embodiment.

[0042] A protein structure-based candidate substance discovery system (1) according to one embodiment can be implemented as a cloud platform that provides users with functions or services necessary for protein structure-based candidate substance discovery in the form of a web service. Specifically, the protein structure-based candidate substance discovery system (1) can provide users with various functions or services, such as, for example, enabling a biologist with a specific idea for new drug development to utilize an in silico screening method, detecting and removing errors (or mistakes) in a protein structure file without knowledge of other fields, efficiently detecting and providing users with an Enzymatically Active Pocket for Docking Calculation (EAPDC) for docking from a protein structure, providing users with a real-time ranking of candidate substances based on docking binding energy while a long-time docking simulation is being performed, increasing reliability by predicting docking binding energy in two stages, or enabling verification of discovered candidate substances through linkage with a verification company.

[0043] The protein structure-based candidate discovery system (1) can provide the same functions or services to users in various environments using a web interface. Specifically, for example, some users may receive services from the protein structure-based candidate discovery system (1) using a mobile device such as a smart phone or tablet computer running a mobile operating system, other users may receive services from the protein structure-based candidate discovery system (1) using a laptop computer running a Windows operating system, and still other users may receive services from the protein structure-based candidate discovery system (1) using a desktop computer running a Linux operating system. In other words, the protein structure-based candidate discovery system (1) is implemented as a cloud platform as a web service, allowing users in different environments to use various functions or services required for calculating in silico candidate substances using an artificial intelligence neural network and conducting subsequent experiments on the candidate substances, such as preclinical experiments, thereby enhancing compatibility and user convenience, and resolving various issues that required improvement in the in silico calculations previously performed through a terminal on Linux.

[0044] Referring to FIG. 1, a protein structure-based candidate material discovery system (1) according to one embodiment may include a project management module (10), a simulation management module (12), a simulation setting module (14), and a simulation flow management module (16).

[0045] The project management module (10) can create a project to which a series of tasks can be added for the discovery of candidate substances based on protein structures. Furthermore, the project management module (10) can display the created project to a user via a display device of a computing device on which the protein structure-based candidate substance discovery system (1) operates.

[0046] In some embodiments, the project management module (10) can perform project name encryption. For a project in which project name encryption is enabled, the project management module (10) can display the original, unencrypted project name to the user who created the project and users with permission to use the project, and display the encrypted project name to users who do not have permission to use the project. There is a concern that the project name may contain keywords related to protein structure-based candidate discovery that must be kept secure. By preventing the project name from being exposed as is to users who are not participating in the same project on the cloud platform, the security of the protein structure-based candidate substance discovery system (1) used by multiple users can be enhanced. Project name encryption can be enabled or disabled by changing the option settings not only when creating a project, but also after the project has been created.

[0047] In some embodiments, the project management module (10) may support project membership via invitation codes. For example, if a user creates a project, the user can send the invitation code to another user, and the other user who receives the invitation code can join the project by entering the invitation code. In other words, users can participate in projects created by others via invitation codes. Accordingly, users with diverse and diverse fields of knowledge can participate in a single project on a cloud platform to conduct protein structure-based candidate material discovery. Furthermore, the project management module (10) may support setting permissions for project members. For example, the project management module (10) may grant administrator permissions to specific members. Furthermore, the project management module (10) may also support users who are participating in a project to leave the project.

[0048] The simulation management module (12) can create a user-defined simulation on a project created by the project management module (10). A single simulation can include multiple task modules with detailed functions for performing protein structure-based candidate discovery, and a single project can include multiple simulations. After creating a simulation, the user can upload protein structure data for performing the multiple task modules.

[0049] The simulation setting module (14) can display a simulation setting area (140) to the user through the display device of the computing device on which the protein structure-based candidate substance discovery system (1) is driven when the user selects one or more simulations managed by the simulation management module (12). In the simulation setting area (140), the user can set the simulation flow so that the simulation desired by the user can be performed by arranging and connecting multiple work modules based on graph computing. In particular, the simulation setting area (140) can be laid out so that the user can see at a glance the function of uploading protein structure data and the work modules that perform detailed work on candidate substance discovery based on the uploaded protein structure data. This configuration can improve the problem that, in the past, detailed work on protein structure-based candidate substance discovery was provided as separate tools with low compatibility, making data sharing and consistent data management difficult. The simulation setting area (140) can include a protein structure data input area (142), a work module selection area (144), and a canvas area (146).

[0050] The protein structure data input area (142) may include one or more objects that can be dragged and dropped onto the canvas area (146). For example, the protein structure data input area (142) may include first to fourth objects. The first object may be converted into a first node by being dragged and dropped onto the canvas area (146), and the first node may receive protein structure data in the form of a PDB (Protein Data Bank) file from the user. The second object may be converted into a second node by being dragged and dropped onto the canvas area (146), and the second node may receive protein structure data in the form of a PDB code from the user. The third object may be converted into a third node by being dragged and dropped onto the canvas area (146), and the third node may receive protein structure data in the form of a protein sequence file from the user. The fourth object can be converted into a fourth node by dragging and dropping it into the canvas area (146), and the fourth node can receive protein structure data in the form of a protein sequence from the user.

[0051] The task module selection area (144) may also include one or more objects that can be dragged and dropped onto the canvas area (146). For example, the task module selection area (144) may include a fifth object to an eighth object. The fifth object may be converted into a fifth node by being dragged and dropped onto the canvas area (146), and the fifth node may perform a task of finding an optimal docking site in a target protein structure. In particular, the fifth node may use an artificial intelligence language model based on Natural Language Processing (NLP) to automatically find an active site in a target protein and generate an optimal docking grid box. In addition, the fifth node may automatically correct various errors that may exist in a protein structure file (i.e., a PDB file). That is, the fifth node can automatically detect and remove anisotropic B-factors from PDB files, detect alternative conformations in the amino acid residue field and correct them to non-alternative conformations, and detect unusual amino acids in the amino acid residue field and correct them to 20 types of non-specific amino acids. The sixth entity can be converted into the sixth node by dragging and dropping it into the canvas area (146), and the sixth node can predict amino acid sequences into protein tertiary structures. The seventh entity can be converted into the seventh node by dragging and dropping it into the canvas area (146), and the seventh node can analyze the binding energy (e.g., in kcal / mol units) of the target protein and the ligand, sort them in order of best results, and provide them to the user.That is, the seventh node can perform grid-based in silico docking based on the input protein structure, determine the chemical pose using LGA (Lamarckian Genetic Algorithm), and calculate the binding energy using the empirical scoring function. The eighth entity can be converted into the eighth node by dragging and dropping it into the canvas area (146), and the eighth node can perform comparative analysis by converting the predicted binding energy between the target protein and the ligand into binding affinity. For example, the eighth node can predict the binding affinity in units of molar concentration, such as fM, pM, nM, μM, mM, or M, using an artificial intelligence model that has learned the protein-ligand structure and Kd / Ki / IC50 values. The predicted concentration units can be displayed right next to the binding affinity values ​​so that a person skilled in the art can intuitively recognize the required values.

[0052] The first to eighth objects can be converted into the first to eighth nodes by dragging and dropping them onto the canvas area (146), respectively. The user can freely arrange the first to eighth nodes onto the canvas area (146) in the desired execution order according to the purpose and environment of the simulation. In addition, the user can set the connection relationship between the first to eighth nodes arranged on the canvas area (146) through edges. The user can create a candidate substance discovery simulation by arranging nodes and connecting edges between nodes. In addition, the user can clearly and easily manage information on protein structure data input by creating, changing, and deleting nodes in a simulation flow whose complexity increases as the number of nodes placed on the canvas area (146) increases and the number of edges between nodes increases.

[0053] In some embodiments, the simulation setting area (140) may further include an external module provision area (148). The external module provision area (148) may include a ninth entity that can be converted into a ninth node having any function provided externally to the protein structure-based candidate substance discovery system (1) by being dragged and dropped into the canvas area (146). Accordingly, functions provided by other systems operating in conjunction with the protein structure-based candidate substance discovery system (1) can be easily included in the simulation by setting a node in the canvas area (146).

[0054] As described above, the user can place and freely move nodes by clicking and dragging a desired object in the protein structure data input area (142), the work module selection area (144), and the external module provision area (148) and dropping it on the canvas area (146). In some embodiments, a node connection shape may be displayed on a node placed on the canvas area (146). While the user clicks and holds a node connection shape displayed on the right side of a node, the color or shape of the node connection shape displayed on the left side of a node that can be connected among other nodes may change. The user cannot establish a connection for a node whose color of the node connection shape displayed on the left side has not changed, and can establish a connection only for a node whose color of the node connection shape displayed on the left side has changed. Accordingly, the user can prevent the creation of an incorrect simulation flow by changing the color of the node connection shape without having to be aware of whether a causal relationship can be established between each node.

[0055] Information for determining whether a connection between nodes is possible can be managed by the simulation flow management module (16). The simulation flow management module (16) can manage information on nodes that can precede or follow in the simulation setup, and can use a separate data structure such as metadata as needed. In addition, the simulation flow management module (16) can also update information, such as by reflecting changes in information on nodes that can precede or follow in the metadata.

[0056] When a user wants to make a connection to another node whose color has changed in the node connection shape displayed on the left side while clicking and holding the node connection shape displayed on the right side of a node, the user can make a connection by clicking the node connection shape of a node and then clicking the node connection shape of another node, or by clicking and holding the node connection shape of a node and dragging it to the node connection shape of another node to make a connection. When the connection is completed, a connection line will appear between the nodes, and the connection between the nodes can be disconnected by clicking the X displayed on the connection line.

[0057] Objects included in the task module selection area (144) can be converted into nodes on the canvas area (146), and a run button can be created within the node. The user can click the run button to perform the task of the corresponding node. Before the task begins, the number of tokens required to perform the task is displayed, and upon user confirmation, the tokens are deducted and the task can be executed. When the task is completed, the run button changes to a result button, and a download button can be additionally created. The user can view the task execution results by clicking the result button, and can download the task execution results by clicking the download button.

[0058] In this way, by establishing the connection between nodes with individual functions in the discovery of candidate substances, a simulation flow suitable for optimal compound development can be created. Furthermore, even when functions are added internally or externally to the protein structure-based candidate substance discovery system (1), nodes corresponding to the added functions can be created, easily establishing connections with existing nodes.

[0059] FIGS. 2 to 6 are diagrams showing exemplary screens of a protein structure-based candidate substance discovery system according to one embodiment.

[0060] Referring to FIG. 2, the simulation setting area (140) may include a protein structure data input area (142), a work module selection area (144), a canvas area (146), and an external module provision area (148). As illustrated, the protein structure data input area (142) may include one or more objects related to a function of uploading protein structure data, the work module selection area (144) may include one or more objects related to detailed work related to candidate substance discovery performed based on the uploaded protein structure data, and the external module provision area (148) may include one or more objects related to any function provided externally. Such objects may be converted into nodes by being dragged and dropped by a user into the canvas area (146), and the nodes may be connected by edges to form a graph, which may represent the flow of the simulation. Of course, the number or types of objects included in the work module selection area (144), canvas area (146), and external module provision area (148) shown in the drawings are examples for the purpose of explaining the implementations, and the scope of the present invention is not limited to those shown.

[0061] Referring to FIG. 3, the protein structure data input area (142) may include a first entity (1420), a second entity (1421), a third entity (1422), and a fourth entity (1423). The first entity (1420), indicated as “PDB File Upload,” is related to a function of receiving protein structure data in the form of a PDB file, and the second entity (1421), indicated as “PDB Code Input,” is related to a function of receiving protein structure data in the form of a PDB code. The third entity (1422), indicated as “Protein Sequence File (Fasta),” is related to a function of receiving protein structure data in the form of a protein sequence file, and the fourth entity (1423), indicated as “Protein Sequence (File),” is related to a function of receiving protein structure data in the form of a protein sequence.

[0062] Meanwhile, the work module selection area (144) may include a fifth entity (1440), a sixth entity (1441), a seventh entity (1442), and an eighth entity (1443). The fifth entity (1440), indicated as "PocketFinder," is related to a function of automatically finding an optimal docking site, and the sixth entity (1441), indicated as "CaliciFold," is related to a function of predicting an amino acid sequence into a protein tertiary structure. The seventh entity (1442), indicated as "AI-Dock," is related to a function of automatically sorting the amino acid sequences into the best order by analyzing the binding energy between the target protein and the ligand, and the eighth entity (1443), indicated as "DeepCalici," is related to a function of converting the predicted binding energy between the target protein and the ligand into binding affinity and performing a comparative analysis.

[0063] In some embodiments, the fifth entity (1440) may detect and remove anisotropic B-factors from PDB files among protein structure data files input by the user, detect alternative conformations in the amino acid residue field and correct them to non-alternative conformations, and detect unusual amino acids in the amino acid residue field and correct them to 20 types of non-specific amino acids. For example, if a docking simulation is performed without removing anisotropic B-factors from the PDB file, an error may occur due to failure to recognize the PDB file format or failure to read the PDB file, and if a docking simulation is performed with alternative conformations or unusual amino acids in the amino acid residue field, an error may occur due to an unknown amino acid, which may result in a decrease in accuracy in performing the in silico screening method or an increase in the failure rate in discovering candidate substances. By automatically handling such error-causing factors, the user can avoid inefficiencies and inaccuracies that may arise from manually modifying PDB files, and can omit collaboration with structural biologists. In addition, the PDB file preprocessing process can be automatically handled internally without the user being aware of it, providing an environment where the user can focus solely on discovering candidate substances. In addition, in some embodiments, the fifth entity (1440) can perform corrections for missing residues in the protein structure of the PDB file. Specifically, the missing residues can be detected by examining the gaps between residues in the protein structure of the PDB file, and when the missing residues are found, a protein amino acid sequence suitable for completing the missing residues can be obtained through a sequence database search, and the missing residues can be automatically completed in the protein amino acid sequence thus obtained.Accordingly, subsequent work can be performed based on an error-free protein structure file from which errors that may occur during simulation have been removed.

[0064] The fifth entity (1440) can detect an Enzymatically Active Pocket for Docking Calculation (EAPDC) from a protein structure file to determine a docking calculation site. Specifically, the fifth entity (1440) can predict a docking site (i.e., EAPDC) in a target protein structure using an artificial intelligence language model. Specifically, the fifth entity (1440) calculates depth values ​​of a pocket based on a Solvent Accessible Surface (SAS) of a target protein surface, generates a Gradient Class Activation Map for amino acids that contributed in the process of predicting the activity of the target protein, and, considering the depth values ​​of the pocket and the values ​​for amino acids with high contributions in the Gradient Class Activation Map, can determine a site that has a high influence on the activity of the target protein as a docking calculation site. Here, the gradient class activation map may be extracted from a GCN (Graph Convolutional Network) trained using EC numbers (Enzyme Commission numbers) or GO numbers (Gene Ontology numbers), which are implemented as a natural language processing model embedding layer. Furthermore, the natural language processing model implemented as an embedding layer may be a Transformer natural language processing model.

[0065] Meanwhile, the external module provision area (148) includes a ninth entity (1480), and the ninth entity (1480) indicated as “CRO-Order” is related to the function of transmitting a verification request for a candidate substance desired by the user to the verification company server.

[0066] As illustrated in FIG. 3, for example, a user can drag and drop a first object (1420) marked as “Upload PDB File” into the canvas area (146), whereby it can be converted into a node (N31). The node (N31) can include a button that enables the function of inputting protein structure data in the form of a PDB file. In addition, the node (N31) can display related information such as the identifier of the corresponding node, the task execution status, etc. In addition, the node (N31) can include a button that can delete itself. As illustrated with the node (N31), the second object (1421) to the ninth object (1480) can be dragged and dropped into the canvas area (146) to be converted into nodes that display their own buttons, information, etc.

[0067] As illustrated in FIG. 4, a user can drag and drop a second object (1421) indicated as “Enter PDB Code” into the canvas area (146), whereby it can be converted into a node (N41). The node (N41) can include a button that enables a function of inputting protein structure data in the form of a PDB code. Subsequently, the user can drag and drop a fifth object (1440) indicated as “PocketFinder” into the canvas area (146), whereby it can be converted into a node (N42). The node (N42) can include a button that enables a function of automatically finding an optimal docking site.

[0068] As illustrated in FIG. 5, a node connection shape may be displayed on a node placed in the canvas area (146). Specifically, a node connection shape (CS1) may be displayed on the right side of a node (N51), and a node connection shape (CS2) may be displayed on the left side of a node (N52). While the user is clicking on the node connection shape (CS1) displayed on the right side of a node (N51), the color or shape of the node connection shape (CS1) displayed on the left side of a node (N52) that can be connected among other nodes may be changed. The user cannot establish a connection for a node whose color of the node connection shape displayed on the left side has not changed, and can establish a connection only for a node whose color of the node connection shape displayed on the left side has changed. While the user is clicking on the node connection shape (CS1) displayed on the right side of a node (N51), the color or shape of the node connection shape (CS1) displayed on the left side of a node (N52) that can be connected among other nodes can be changed based on information about preceding or succeeding nodes provided from the simulation flow management module (16).

[0069] As illustrated in FIG. 6, when you want to make a connection to another node (N62) whose color has changed in the node connection shape (CS2) displayed on the left side while clicking on the node connection shape (CS1) displayed on the right side of a node (N61), you can make a connection by clicking on the node connection shape (CS1) of a node (N61) and then clicking on the node connection shape (CS2) of another node (N62), or you can make a connection by clicking and dragging the node connection shape (CS1) of a node (N61) to the node connection shape (CS2) of another node (N62). When the connection is completed, a connection line is displayed between the nodes, and you can release the connection between the nodes by clicking on the X displayed on the connection line.

[0070] Figure 7 is a diagram showing an example of a simulation setting of a protein structure-based candidate material discovery system according to one embodiment.

[0071] Referring to FIG. 7, in an example of generating a simulation flow, nodes (N71) to (N74) are illustrated. The node connection shape on the right side of the node (N71) that inputs protein structure data in the form of PDB code and the node connection shape on the left side of the node (N72) that automatically finds an optimal docking site are connected by an edge, the node connection shape on the right side of the node (N72) and the node connection shape on the left side of the node (N73) that analyzes and automatically aligns the binding energy of the target protein and the ligand are connected by an edge, and the node connection shape on the right side of the node (N73) and the node connection shape on the left side of the node (N74) that converts the predicted binding energy between the target protein and the ligand into binding affinity and performs comparative analysis are connected by an edge. As explained above, the user cannot establish a connection for a node whose color of the node connection shape displayed on the left side has not changed among other nodes, and can establish a connection only for a node whose color of the node connection shape displayed on the left side has changed. Therefore, the user does not have to consider whether nodes can precede or follow each other when discovering candidate materials, which increases convenience.

[0072] In some embodiments, information about possible preceding or succeeding nodes may be predetermined as follows.

[0073] Preceding NodesNodesFollowing NodesUpload PDB FilePocketFinderAI-DockInput PDB CodeProtein Sequence File (Fasta)Protein Sequence (Text)CaliciFoldProtein Sequence File (Fasta)CaliciFoldPocketFinderProtein Sequence (Text)PocketFinderAI-DockDeepCaliciAI-DockDeepCalici-

[0074] Nodes that can precede a node that automatically finds an optimal docking site ("PocketFinder") may include a node that receives protein structure data in the form of a PDB file ("Upload PDB File"), a node that receives protein structure data in the form of a PDB code ("Enter PDB Code"), a node that receives protein structure data in the form of a protein sequence file ("Protein Sequence File (Fasta)"), a node that receives protein structure data in the form of a protein sequence ("Protein Sequence (Text)"), and a node that predicts an amino acid sequence as a protein 3-dimensional structure ("CaliciFold"), and nodes that can follow it may include a node that analyzes the binding energy of a target protein and a ligand and automatically aligns them ("AI-Dock"). Nodes that can precede a node that automatically finds an optimal docking site ("CaliciFold") may include a node that receives protein structure data in the form of a protein sequence file ("Protein Sequence File (Fasta)"), a node that receives protein structure data in the form of a protein sequence ("Protein Sequence (Text)"), and nodes that can follow it may include a node that automatically finds an optimal docking site. May contain a node ("PocketFinder").

[0075] A node that can precede a node that automatically aligns target proteins and ligands by analyzing the binding energy of the target protein and ligand ("AI-Dock") may include a node that automatically finds the optimal docking site ("PocketFinder"), and a node that can follow a node that converts the predicted binding energy between the target protein and ligand into binding affinity and performs comparative analysis ("DeepCalici").

[0076] A node that can precede a node that performs comparative analysis by converting predicted binding energy between a target protein and a ligand into binding affinity ("DeepCalici") may include a node that automatically aligns the target protein and ligand by analyzing the binding energy of the target protein and ligand ("AI-Dock").

[0077] Information on such preceding or succeeding nodes can be managed by the simulation flow management module (16), and a separate data structure, such as metadata, can be used as needed. Furthermore, the simulation flow management module (16) can also update information, such as by reflecting changes in the information on preceding or succeeding nodes in the metadata. In this way, by establishing the connection relationship between nodes with each function in the discovery of candidate substances, a simulation flow suitable for developing optimal compounds can be created.

[0078] Conventional simulation methods require numerous attempts in various ways to discover protein structure-based candidates. Complex simulations require significant effort, time, and cost, and satisfactory simulation setup is difficult. The graph computing-based simulation setup method described in the examples improves upon the conventional method. As exemplified, it allows for the intuitive and easy creation and management of complex simulation flows, while also offering flexibility and convenience for easy modification. Furthermore, it enables the practical execution of highly complex and intricate simulation flows. Furthermore, it provides optimized functions and a user interface for candidate discovery. Previously, detailed tasks related to candidate discovery were provided as separate, incompatible tools, making data sharing and consistent data management difficult. Furthermore, the simulation flow can be easily managed by creating, modifying, and deleting nodes.

[0079] FIGS. 8 to 10 are drawings illustrating the operation of a protein structure-based candidate material discovery system according to one embodiment.

[0080] Referring to FIG. 8, a protein structure-based candidate substance discovery system according to one embodiment can display a first screen (30) that provides a plurality of ligands to be docked to a target protein.

[0081] In the present embodiment, the simulation setting module (14) of the protein structure-based candidate substance discovery system displays a simulation setting area (140) to the user, and as described above, the simulation setting area (140) may include a work module selection area (144) and a canvas area (146). The simulation setting module (14) may set a simulation flow for a simulation that is set based on the user's dragging and dropping of a plurality of work modules onto the canvas area (146) and arranging and connecting them based on graph computing. In particular, in the present embodiment, the work module selection area (144) may include a first object and a second object. The first object may be changed into a first node that calculates an expected binding energy according to docking of a target protein and a ligand by being dragged and dropped onto the canvas area (146), and the second object may be changed into a second node that predicts a binding affinity corresponding to a protein-ligand binding pose used in the calculation of the binding energy calculated in the first node by being dragged and dropped onto the canvas area (146).

[0082] As described above, the first node and the second node may each include a run button. A user may click the run button displayed on the first node to perform a task of calculating the expected binding energy based on docking between the target protein and the ligand, and click the run button displayed on the second node to perform a task of predicting the binding affinity corresponding to the protein-ligand binding pose used in the binding energy calculation calculated in the first node.

[0083] When the execution button displayed on the first node is clicked, the first screen (30) may be displayed. The user can then set the necessary items to perform a task through the first screen (30). The first screen (30) may include a plurality of user interface elements. In this specification, the term "user interface elements" is used as a concept that includes all visual elements that a user interacts with, such as buttons, labels, text boxes, images, sliders, and drop-down menus, and in contrast, the term "user interface controls" is used as a concept that indicates elements that receive user input and transmit commands to an application, such as buttons, check boxes, radio buttons, and switches.

[0084] The first screen (30) may include a plurality of user interface elements, including a first user interface element (301). The first user interface element (301) may be used to set the number of results to be displayed after the execution of the task of the first node is completed. For example, when a user inputs a value of '500' into the first user interface element (301), the top 500 binding energies expected according to docking of the target protein and the ligand may be displayed after the execution of the first task is completed.

[0085] Among the plurality of user interface elements of the first screen (30), the second user interface element (302) may be used to set a ligand to be docked with the target protein. For example, the second user interface element (302) may include 'random' and 'upload' as its values, and if the user selects 'random', the first task may perform docking with the target protein based on a ligand library provided by the protein structure-based candidate substance discovery system. Alternatively, if the user selects 'upload', the user directly uploads data on the ligand, and the first task may perform docking with the target protein based on the ligand data uploaded by the user.

[0086] Among the plurality of user interface elements of the first screen (30), the third user interface element (303) can be used to set the type and number of ligands to be docked with the target protein. For example, when the user inputs 'FDA', 'All', and '2115' into the third user interface element (303), docking can be performed for all 2115 ligands in the FDA-approved drug library. Or, as another example, when the user inputs 'MCULE', 'In Stock', and '2000' into the third user interface element (300), docking can be performed for 2000 In Stock compounds that are in stock and ready for immediate delivery from the Mcule library.

[0087] Among the plurality of user interface elements of the first screen (30), the fourth user interface element (304) may be used to set the number of GPUs to be used for the task of calculating the expected binding energy according to the docking of the target protein and the ligand performed by the first node. For example, if the user inputs '1' in the fourth user interface element (304), one GPU thread may be used to perform the task, and if '2' is input, two GPU threads may be used to perform the task.

[0088] Among the plurality of user interface elements of the first screen (30), the fifth user interface element (305) may display the number of tokens required to execute a task of calculating the expected binding energy according to the docking of the target protein and the ligand performed by the first node. The protein structure-based candidate material discovery system is implemented so that the tokens that the user must pay for the detailed tasks of the simulation for candidate discovery can be deducted, or the tokens can be recharged when the user pays the cost through a coupon or various payment methods. The amount of tokens to be deducted can be determined by considering various factors of the detailed task, such as the type of detailed task, the workload of the detailed task, and the difficulty of the detailed task. Data regarding the deduction criteria for the token amount can be stored in a form readable by the computing device on a storage medium or in the cloud that the computing device can access. The user can use tokens based on the deduction criteria determined according to the detailed task to be performed. In particular, as described above, in the discovery of candidate substances, a simulation flow suitable for the development of optimal compounds can be created by setting the connection relationship of nodes with each function, and the user can use the corresponding node by paying a token that is differentially applied to each node, and the corresponding node can pay only the amount of tokens needed. For example, the fifth user interface element (305) can indicate that 2,115 tokens, which is the number of ligands in the case of an FDA-approved drug library, are required for the simulation task performed by the first node. If the value set in the fourth user interface element (304) is changed from '1' to '2', two GPU threads are allocated to perform the simulation task for 2,115 ligands, which can reduce the task time by half while paying twice the cost.

[0089] Among the plurality of user interface elements of the first screen (30), the sixth user interface element (306) can submit the settings made through the first screen (30), sequentially perform molecular docking on a plurality of ligands according to the settings input by the user through the first screen (30), and calculate the expected binding energy.

[0090] Here, the first user interface element (301), the second user interface element (302), the third user interface element (303), the fourth user interface element (304), and the sixth user interface element (306) may be treated as user interface controls differently from the fifth user interface element (305).

[0091] Referring to FIG. 9, a protein structure-based candidate substance discovery system according to one embodiment may display a second screen (31) that provides the user with the predicted binding energy as a result based on the docking of the target protein and the ligand. As described above, the user can check the results sorted in order of binding energy through the second screen (31) by clicking the result button after the execution button displayed on the first node has changed to a result button. In some embodiments, a download button may be additionally created on the first node, and the user may also download the results sorted in order of binding energy by clicking the download button.

[0092] The second screen (31) may include a plurality of user interface elements. Among the plurality of user interface elements of the second screen (31), the first user interface element (311) may be a user interface control for user selection, for example, a check box. In some embodiments, for convenience when a user selects multiple rows, a button may be additionally displayed to allow the user to select all rows on the displayed page at once (“Select All in Page”), or a button may be additionally displayed to allow the user to select all results at once (“Select All”).

[0093] Among the plurality of user interface elements of the second screen (31), the second user interface element (312) can display the names of the plurality of ligands for which molecular docking was performed.

[0094] Among the plurality of user interface elements of the second screen (31), the third user interface element (313) can display the binding energy predicted through docking of the target protein and the plurality of ligands. Here, the third user interface element (313) can display the value of the binding energy as a value according to a unit representing the amount of energy. For example, here, the third user interface element (313) can display the value of the binding energy in units of Kcal / mol.

[0095] Among the plurality of user interface elements of the second screen (31), the fourth user interface element (314) can display the structure of a substance resulting from molecular docking in the form of a string. For example, the fourth user interface element (314) can display the structure of a substance resulting from molecular docking in the form of a string according to the Simplified Molecular Input Line Entry System (SMILES) notation. In the SMILES notation, for example, atoms can be represented by element symbols, and bonds can be represented by specific characters such as '=', '#', etc. In addition, branches within a molecule can be represented using parentheses, rings can be represented by numbers, or chirality can be represented by specific symbols such as '@'.

[0096] The second screen (31) displays the results as a first list consisting of a plurality of rows, one row (315) of which may include user interface controls for user selection, names of a plurality of ligands, and values ​​of the calculated binding energies.

[0097] Among the plurality of user interface elements of the second screen (31), the fifth user interface element (316) can submit information on rows selected through the first user interface element (311) in the first list of the second screen (31), and cause binding affinity corresponding to the protein ligand binding pose used in binding energy calculation to be calculated.

[0098] Here, the first user interface element (311) and the fifth user interface element (316) may be treated as user interface controls differently from the second user interface element (312), the third user interface element (313), and the fourth user interface element (314).

[0099] Referring to FIG. 10, a protein structure-based candidate substance discovery system according to one embodiment may display a third screen (32) that provides the user with predicted binding affinities corresponding to binding energies corresponding to rows selected by the user as results. The user may click the fifth user interface element (316) on the second screen (32) to check the results sorted in order of binding energy through the third screen (32). Alternatively, when moving to a screen where nodes are displayed from the second screen (32), the user may click the result button displayed on the second node to check the results sorted in order of binding affinity through the third screen (32). In some embodiments, a download button may be additionally created on the second node, and the user may also click the download button to download the results sorted in order of binding affinity.

[0100] The third screen (32) may include a plurality of user interface elements. Among the plurality of user interface elements of the third screen (32), the first user interface element (321) may display a sequence sorted according to binding affinity.

[0101] Among the multiple user interface elements of the third screen (32), the second user interface element (322) can display the names of multiple ligands for which molecular docking was performed. Specifically, among those displayed by the second user interface element (312) of the second screen (31), only those selected by the user through the first user interface element (311) can be displayed.

[0102] Among the multiple user interface elements of the third screen (32), the third user interface element (323) can display the binding energy predicted through docking of the target protein and multiple ligands. Specifically, among those displayed by the third user interface element (313) of the second screen (31), only those selected by the user through the first user interface element (311) can be displayed.

[0103] Among the plurality of user interface elements of the third screen (32), the fourth user interface element (324) can display a binding affinity corresponding to a protein-ligand binding pose used for binding energy calculation. Here, the fourth user interface element (324) can display the value of the binding affinity as a value according to the unit of molar concentration. For example, here, the fourth user interface element (324) can display the value of the binding affinity in units of fM, pM, nM, μM, mM, or M.

[0104] Among the multiple user interface elements of the third screen (32), the fifth user interface element (315) can display the structure of a substance resulting from molecular docking in the form of a string. Specifically, among those displayed by the fourth user interface element (314) of the second screen (31), only those selected by the user through the first user interface element (311) can be displayed.

[0105] The third screen (32) displays the results as a second list consisting of a plurality of rows, and one row (326) may include the names of a plurality of ligands, values ​​of binding energy, and values ​​of binding affinity.

[0106] In this embodiment, the rows forming the second list may be displayed on a third screen, sorted in descending or ascending order based on the binding affinity value. Of course, it may also be possible to sort in descending or ascending order based on the binding energy value on the same third screen.

[0107] In this way, by predicting the binding affinity between a target protein and a ligand and presenting the binding energy and binding affinity in parallel, i.e., as adjacent columns, on a single screen, the user can intuitively understand the results and easily calculate the molecular weight of a candidate substance, such as a new drug candidate, and the volume of the medium, without separate additional experiments, thereby directly applying the results to clinical trials such as cell experiments. Therefore, the time and cost consumed in determining the experimental concentration of a candidate substance can be significantly reduced.

[0108] FIG. 11 is a diagram illustrating an operation method of a protein structure-based candidate substance discovery system according to one embodiment.

[0109] According to one embodiment, a method for operating a protein structure-based candidate substance discovery system is provided as a method for operating a protein structure-based candidate substance discovery system implemented as a cloud platform that provides users with functions or services necessary for candidate substance discovery in the form of a web service, the method including: displaying a first screen for receiving a plurality of ligands to be docked to a target protein; sequentially performing molecular docking on the plurality of ligands according to settings input by a user through the first screen and calculating an expected binding energy; displaying a first list composed of rows including user interface controls for user selection, names of the plurality of ligands, and values ​​of the calculated binding energies on a second screen; predicting binding affinities corresponding to protein-ligand binding poses used for binding energy calculation for rows selected by the user using the user interface controls from the first list on the second screen; and displaying a second list composed of rows including names of the plurality of ligands, values ​​of the binding energies, and values ​​of the binding affinities on a third screen. For more detailed information on how the protein structure-based candidate substance discovery system works, please refer to the previously described contents with reference to FIGS. 1 to 10, and thus, redundant description will be omitted here.

[0110] In some embodiments, the binding affinity may be predicted by an artificial intelligence model trained using experimental measurements including at least one of a dissociation constant Kd, an inhibition constant Ki, and a half maximal inhibitory concentration IC50 as training data. Here, the artificial intelligence model may be trained using the experimental measurements and binding structure data of the target protein and the ligand as training data. In some embodiments, the artificial intelligence model may include a convolutional neural network (CNN) model or a residual network 3D (ResNet 3D) model.

[0111] Referring to FIG. 11, learning of an artificial intelligence model can be performed according to steps (S1101) to (S1108), and prediction of binding affinity based on the artificial intelligence model can be performed according to steps (S1109) to (S1111).

[0112] In step (S1101), the binding structure of a protein and a ligand and experimentally measured values ​​can be input as learning data. The binding structure of a protein and a ligand can be a three-dimensional structure or PDB data, and the experimentally measured value can be at least one of Kd, Ki, and IC50.

[0113] In step (S1102), outliers present in the training data can be removed. Outliers are values ​​that are physically unlikely to exist due to their characteristics, and are likely to correspond to measurement errors or structural errors in proteins. For example, if outliers exist in the PDBBind data set, a histogram-based outlier removal technique can be used to remove them.

[0114] In step (S1103), missing residues are detected in the protein structure and ligand, and if missing residues are found, a protein amino acid sequence suitable for completing the missing residues is obtained through a sequence database search, and the missing residues can be automatically completed in the protein amino acid sequence thus obtained. Protein structures that cannot be modified can be deleted.

[0115] In step (S1104), data sets can be separated to prevent similar structures from being mixed into the validation or test data sets. For example, when predicting a three-dimensional protein structure, data sets can be separated based on the Template Modeling Score (TM Score), an indicator that evaluates the similarity of the predicted structure to the experimentally determined actual structure.

[0116] In step (S1105), learning for an artificial intelligence model can be performed using learning data prepared through previous steps.

[0117] In steps (S1106) to (S1108), feature importance is calculated based on the SHAP (SHapley Additive exPlanations) value, which numerically evaluates the contribution of each input variable to each prediction, to explain how each input feature of the artificial intelligence model contributes to the final prediction, and feature engineering, which adds or removes features through visual inspection using a protein structure analysis tool, can be performed.

[0118] Meanwhile, in step (S1109), a three-dimensional structural pose of a protein to be predicted and a candidate substance combined is input into an artificial intelligence model whose learning has been completed, and in step (S1110), a binding affinity based on molar concentration can be obtained as a predicted value of the artificial intelligence model so that a user with only biological knowledge can dissolve the protein in a drug. In step (S1111), the binding affinity can be displayed to the user on the screen described above with reference to FIG. 10.

[0119] Fig. 12 is a diagram showing an example of an implementation of an artificial intelligence model of a protein structure-based candidate substance discovery system according to one embodiment.

[0120] Referring to FIG. 12, an artificial intelligence model for predicting binding affinity in a protein structure-based candidate substance discovery system according to one embodiment may include a convolutional layer and a dense layer. The convolutional layer may include filters encoding patterns to be identified to predict binding affinity between a target protein and a ligand. Meanwhile, the dense layer may integrate features extracted through the convolutional layer.

[0121] In some embodiments, the artificial intelligence model may be a deep 4D convolutional neural network, specifically a deep 4D convolutional neural network having a single output neuron for predicting binding affinity between a target protein and a ligand.

[0122] The convolution layer can identify patterns encoded by the filters of the convolution layers and generate a feature map that highlights the spatial occurrences of each pattern in the data. Specifically, the convolution layer can include three convolution layers, each having 64, 128, and 256 filters, respectively, and the result of the final convolution layer can be flattened and used as input to the dense layer.

[0123] The dense layer section includes three dense layers, each with 1,000, 500, and 200 neurons, respectively. Dropout with a drop probability of 0.5 can be used for all dense layers. Additionally, the L2 regularization technique with a parameter (lambda) value of 0.001 can be used.

[0124] Meanwhile, ReLU (Rectified Linear Unit) can be adopted as an activation function in both the convolutional layer and the dense layer.

[0125] FIG. 13 is a block diagram illustrating a computing device according to one embodiment.

[0126] Referring to FIG. 13, a protein structure-based candidate material discovery system according to embodiments can be implemented using a computing device (50).

[0127] The computing device (50) may include at least one of a processor (501), a memory (502), a storage device (503), a display device (504) communicating via a bus (509), a network interface device (505) providing a connection to a network (40) for communication with other entities, and an input / output interface device (506) providing a user input interface or a user output interface. Of course, the computing device (50) may additionally include any electronic device necessary to implement the technical ideas described herein, although not illustrated in FIG. 13.

[0128] The processor (501) may be implemented in various types such as an Application Processor (AP), a Central Processing Unit (CPU), a Graphic Processing Unit (GPU), a Neural Processing Unit (NPU), etc., and may be any electronic device that executes a program or command stored in a memory (502) or a storage device (503). In particular, the processor (501) may be configured to implement the functions or methods described above with respect to FIGS. 1 to 12, and in relation to the protein structure-based candidate substance discovery system and its operating method according to embodiments of the present invention, artificial intelligence-specific operations may be processed on a GPU or NPU.

[0129] The memory (502) and the storage device (503) may include various types of volatile or non-volatile storage media. For example, the memory (502) may include a read-only memory (ROM) or a random access memory (RAM), and the memory (502) may be located inside or outside the processor (501) and may be connected to the processor (501) through various known means. Meanwhile, examples of the storage device (503) include a hard disk drive (HDD) or a solid state drive (SSD), and the scope of the present invention is not limited to the elements listed above for explanation.

[0130] The protein structure-based candidate material discovery system and its operating method according to the embodiments can be implemented as a program or software running on a computing device (50), and such a program or software can be stored on a computer-readable medium.

[0131] Meanwhile, the protein structure-based candidate material discovery system and its operation method according to the embodiments may be implemented using the hardware of the computing device (50), or may be implemented as separate hardware that can be electrically connected to the computing device (50).

[0132] According to the embodiments described so far, functions and a user interface optimized for candidate material discovery are provided, and the difficulties in data sharing and consistent data management due to the detailed tasks related to candidate material discovery being provided as separate tools with high incompatibility with each other are improved, and the simulation flow can be easily managed by creating, changing, and deleting nodes.

[0133] In conjunction with this, by predicting and providing the binding affinity between the target protein and the ligand to the user, it helps the user intuitively understand the results and allows the user to easily calculate the molecular weight of a candidate substance, such as a new drug candidate, and the volume of the medium, without additional experiments, allowing direct application to clinical trials such as cell experiments. Therefore, the time and cost consumed in determining the experimental concentration for the required candidate substance can be significantly reduced. In particular, since these experimental concentrations are provided through a cloud platform, users can easily access the experimental concentration values ​​from anywhere with an internet connection, and collaboration with other users is also convenient.

[0134] Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art to which the present invention pertains using the basic concept of the present invention defined in the following claims also fall within the scope of the present invention.

Claims

1. A method of operating a protein structure-based candidate substance discovery system implemented as a cloud platform that provides users with functions or services necessary for protein structure-based candidate substance discovery in the form of a web service. A step of displaying a first screen in which a plurality of ligands for docking to a target protein are provided; A step of sequentially performing molecular docking on the plurality of ligands according to the settings input by the user through the first screen and calculating the expected binding energy; A step of displaying a first list consisting of rows containing names of the plurality of ligands and values ​​of the calculated binding energies on a second screen, a user interface control for user selection; In the first list of the second screen, a step of predicting binding affinity corresponding to the protein ligand binding pose used in the binding energy calculation for rows selected by the user using the user interface control; and A step of displaying a second list consisting of rows including the names of the plurality of ligands, the values ​​of the binding energy, and the values ​​of the binding affinity on a third screen, How a protein structure-based candidate substance discovery system works.

2. In paragraph 1, The value of the above binding energy is expressed as a value according to the unit representing the amount of energy, A method of operation of a protein structure-based candidate substance discovery system, wherein the above binding affinity value is expressed as a value according to the unit of molar concentration.

3. In paragraph 2, The values ​​of the above binding energies are expressed in units of Kcal / mol, A method of operation of a protein structure-based candidate substance discovery system, wherein the values ​​of the above binding affinity are expressed in units of fM, pM, nM, μM, mM or M.

4. In paragraph 3, The step of displaying the second list on the third screen is: An operating method of a protein structure-based candidate substance discovery system, comprising a step of displaying the rows forming the second list on the third screen in descending or ascending order based on the value of the binding affinity.

5. In paragraph 2, The above binding affinity is, An operating method of a protein structure-based candidate substance discovery system, wherein the predicted value is an artificial intelligence model trained using experimental measurements including at least one of a dissociation constant Kd, an inhibition constant Ki, and a half maximal inhibitory concentration IC50 as learning data.

6. In paragraph 5, The above artificial intelligence model is, An operating method of a protein structure-based candidate substance discovery system, wherein the above experimental measurement values ​​and the binding structure data of the target protein and the ligand are learned as the learning data.

7. In paragraph 6, The above artificial intelligence model is, An operating method of a protein structure-based candidate substance discovery system, comprising a convolution layer section including filters encoding patterns to be identified to predict binding affinity between the target protein and the ligand, and a dense layer section integrating features extracted through the convolution layer section.

8. In paragraph 7, The above convolution layer is, A method of operation of a protein structure-based candidate discovery system comprising three convolutional layers having 64, 128 and 256 filters respectively.

9. In paragraph 8, The above dense layer is, A method of operation of a protein structure-based candidate discovery system comprising three dense layers having 1000, 500 and 200 neurons, respectively.

10. In paragraph 5, The above artificial intelligence model is, An operating method of a protein structure-based candidate substance discovery system including a CNN (Convolutional Neural Network) model or a ResNet 3D (Residual Network 3D) model.

11. A protein structure-based candidate substance discovery system implemented as a cloud platform that provides users with functions or services necessary for protein structure-based candidate substance discovery in the form of a web service. A project management module that creates projects that can be added to perform tasks for the discovery of candidate compounds based on protein structure; A simulation management module that creates a simulation desired by the user on the generated project; A simulation setting module that sets a simulation flow for the simulation, which is input from a user, using a simulation setting area including a work module selection area, which includes a first object that is changed into a first node for calculating an expected binding energy according to docking of a target protein and a ligand by being dragged and dropped into a canvas area, and a second object that is changed into a second node for predicting a binding affinity corresponding to a protein-ligand binding pose used for calculating the binding energy calculated in the first node; and In the above simulation flow setting, a simulation flow management module is included that manages information about nodes that can precede or follow. Protein structure-based candidate substance discovery system.

12. In paragraph 11, A protein structure-based candidate substance discovery system, wherein when the execution button displayed on the first node is clicked, a first screen is displayed that provides a plurality of ligands for docking to the target protein.

13. In paragraph 12, The above first node, A protein structure-based candidate substance discovery system that sequentially performs molecular docking on the plurality of ligands according to the settings entered by the user through the first screen and calculates the expected binding energy.

14. In paragraph 13, A protein structure-based candidate substance discovery system, wherein after the calculation of the above binding energy is completed, a first list consisting of rows including user interface controls for user selection, names of the plurality of ligands, and values ​​of the calculated binding energy is displayed on a second screen.

15. In paragraph 14, The second node above, A protein structure-based candidate substance discovery system that predicts binding affinity corresponding to a protein ligand binding pose used in the binding energy calculation for rows selected by a user using the user interface control in the first list of the second screen.

16. In paragraph 15, A protein structure-based candidate substance discovery system, wherein after the prediction of the binding affinity is completed, a second list consisting of rows including the names of the plurality of ligands, the values ​​of the binding energy, and the values ​​of the binding affinity is displayed on a third screen.

17. In paragraph 16, The above second list is displayed on the above third screen, A protein structure-based candidate substance discovery system, comprising rows forming the second list being displayed on the third screen in descending or ascending order based on the value of the binding affinity.

18. In paragraph 11, The values ​​of the above binding energies are expressed in units of Kcal / mol, A protein structure-based candidate discovery system, wherein the values ​​of the above binding affinity are expressed in units of fM, pM, nM, μM, mM or M.

19. In paragraph 11, The above binding affinity is, A protein structure-based candidate substance discovery system, wherein the predicted substance is an artificial intelligence model trained using experimental measurements including at least one of a dissociation constant Kd, an inhibition constant Ki, and an inhibition median value IC50 as learning data.

20. In paragraph 11, The above simulation flow management module is a protein structure-based candidate substance discovery system that manages information about nodes that can precede or follow in the above simulation flow setting through metadata.

Images