Method, apparatus, device and medium for predicting anti-aging drug combination

Abstract

The application relates to the field of artificial intelligence medicine and medical health technology, and proposes a prediction method and device for an anti-aging drug combination, equipment and a medium, wherein the method comprises the following steps: acquiring a personal multi-omics expression profile of a target drug use population, wherein the multi-omics comprises a genome, an RNA group, a protein group, a post-translational modification group and a metabolome; according to the personal multi-omics expression profile and a pre-trained multi-omics expression profile perturbation prediction model, the perturbation parameters of the target multi-omics corresponding to the target drug combination under different doses are predicted; and according to the perturbation parameters of the target multi-omics under each dose of the target drug combination, it is determined whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose. Through the technical scheme, an effective anti-aging drug can be comprehensively, efficiently and accurately found from the perspective of systems biology.

Description

Predictive methods, devices, equipment and media for anti-aging drug combinations Technical Field

[0001] This application relates to the fields of artificial intelligence drugs and medical health technology, and in particular to a method, apparatus, device and medium for predicting anti-aging drug combinations. Background Technology

[0002] Aging has always been a hot topic of concern in the medical community and indeed in all of humanity. In recent years, with the increasing elderly population and the intensifying trend of aging, the incidence of age-related chronic diseases, such as diabetes, heart disease, and Alzheimer's disease, has been rising continuously. This not only seriously threatens the lives and health of the elderly but also imposes a heavy social, medical, and economic burden on countries worldwide. Finding effective anti-aging drugs and delaying aging through drug intervention has become a current research hotspot in the field of anti-aging.

[0003] Drug development is an industry characterized by long cycles, high risks, and large investments, with costs reaching billions of yuan. With the rapid advancements in pharmaceutical research and development, traditional experimental methods for predicting drug activity are no longer sufficient to meet the growing demands of drug development.

[0004] Using artificial intelligence algorithms to assist in drug development has become an important means of predicting drug activity in drug development. However, the current field of anti-aging drug discovery lacks models and methods that fully utilize multi-omics information and individual multi-omics expression profiles for prediction, making it difficult to discover effective anti-aging drugs comprehensively, efficiently, and accurately from a systems biology perspective. Summary of the Invention

[0005] This application provides a method, apparatus, device, and medium for predicting anti-aging drug combinations, aiming to solve the technical problems in related technologies such as the lack of models and methods that fully utilize multi-omics information and personal multi-omics expression profiles for prediction.

[0006] In a first aspect, embodiments of this application provide a method for predicting anti-aging drug combinations, including:

[0007] Obtain individual multi-omics expression profiles of target drug users, wherein the multi-omics includes genomics, RNAiomics, proteomics, post-translational modification genomics, and metabolomics;

[0008] Based on the individual multi-omics expression profile and the perturbation prediction model of the pre-trained multi-omics expression profile, the perturbation parameters of the target multi-omics corresponding to the target drug combination at different doses are predicted.

[0009] Based on the perturbation parameters of the target multi-omics at various doses of the target drug combination, it is determined whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose.

[0010] In one embodiment, optionally, the method further includes:

[0011] To obtain the original multi-omics expression profiles of various cell lines and the multi-omics perturbation information of various drugs at different doses on different human tissue cells;

[0012] Based on the original multi-omics expression profile and the multi-omics perturbation information, a preset adversarial residual cell network model is trained to obtain a perturbation prediction model of the multi-omics expression profile.

[0013] In one embodiment, optionally, a preset prediction model is trained based on the original multi-omics expression profile and the multi-omics perturbation information to obtain a perturbation prediction model of the multi-omics expression profile, including:

[0014] The full N-dimensional multi-omics raw expression spectrum is encoded by an encoder to obtain the first n-dimensional latent vector used to characterize the expression level of the target.

[0015] The cell lines corresponding to the original multi-omics expression profiles are characterized by Se ResNet, and a second n-dimensional hidden vector is generated to characterize the cell lines.

[0016] Drug molecules are encoded using a graph neural network model to obtain drug features and generate a third n-dimensional hidden vector to characterize the drug.

[0017] Based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector, the decoder generates an N-dimensional total feature vector.

[0018] The N-dimensional total feature vector is used as input, and the corresponding multi-omics perturbation information is used as output vector to train the preset prediction model, so as to obtain the perturbation prediction model of the multi-omics expression spectrum.

[0019] In one embodiment, optionally, drug molecules are encoded using a graph neural network model to obtain drug features and generate a third n-dimensional latent vector for characterizing the drug, including:

[0020] Drug molecules are encoded using a graph neural network model to obtain drug features, resulting in an m-dimensional hidden vector.

[0021] Drug dosage is represented by a one-dimensional vector, and drug dosage and drug features are concatenated into an m+1 dimensional latent vector.

[0022] The m+1 dimensional latent vector is transformed into the third n-dimensional latent vector using a multilayer perceptron.

[0023] In one embodiment, optionally, an N-dimensional total feature vector is generated using a decoder based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector, including:

[0024] Add the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector to obtain the total n-dimensional latent vector;

[0025] Based on the total n-dimensional latent vector, the decoder generates an N-dimensional total feature vector.

[0026] In one embodiment, optionally, determining whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the perturbation parameters of the target multi-omics at various doses of the target drug combination includes:

[0027] Obtain the non-drug multi-omics expression profile of the young non-drug group corresponding to the target drug user group;

[0028] Based on the perturbation parameters of the target multi-omics at each dose of the target drug combination, the similarity score between the perturbation multi-omics expression profile of the target drug combination at each dose and the untreated multi-omics expression profile is calculated.

[0029] The similarity score is used to determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose.

[0030] In one embodiment, optionally, determining whether the target drug combination is a potential anti-aging drug combination and the corresponding target dosage based on the similarity score includes:

[0031] The similarity scores of the target drug combinations at various dosages are sorted in descending order.

[0032] Target drug combinations with similarity scores greater than preset scores and their corresponding dosages are identified as potential anti-aging drug combinations and target dosages.

[0033] Secondly, embodiments of this application provide a device for predicting anti-aging drug combinations, comprising:

[0034] The first acquisition module is used to acquire the individual multi-omics expression profile of the target drug user group, wherein the multi-omics includes genomics, RNAiome, proteomics, post-translational modification genome and metabolomics;

[0035] The prediction module is used to predict the perturbation parameters of the target multi-omics corresponding to different doses of the target drug combination based on the individual multi-omics expression profile and the perturbation prediction model of the pre-trained multi-omics expression profile.

[0036] The determination module is used to determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the perturbation parameters of the target multi-omics at each dose of the target drug combination.

[0037] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for predicting the combination of anti-aging drugs.

[0038] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described method for predicting anti-aging drug combinations.

[0039] The above-mentioned methods, devices, equipment, and media for predicting anti-aging drug combinations involve obtaining the individual multi-omics expression profile of the target drug user group. This multi-omics profile includes the genomics, RNAitomy, proteomics, post-translational modification genome, and metabolomics. Based on the individual multi-omics expression profile and a pre-trained perturbation prediction model of the multi-omics expression profile, the perturbation parameters of the target multi-omics profile corresponding to the target drug combination at different doses are predicted. Based on the perturbation parameters of the target multi-omics profile at each dose of the target drug combination, it is determined whether the target drug combination is a potential anti-aging drug combination and its corresponding target dose. In this invention, by using a pre-trained perturbation prediction model of the multi-omics expression profile, and fully utilizing multi-omics information and individual multi-omics expression profiles for prediction, effective anti-aging drugs can be comprehensively, efficiently, and accurately discovered from a systems biology perspective. This allows for the comprehensive regulation of drug combinations across all pathways and targets in the human body, thereby improving the occurrence and development of age-related diseases. Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 shows a schematic flowchart of a method for predicting an anti-aging drug combination according to an embodiment of this application.

[0042] Figure 2 shows a schematic flowchart of a method for predicting an anti-aging drug combination according to another embodiment of this application.

[0043] Figure 3 shows a schematic flowchart of step S202 in a method for predicting an anti-aging drug combination according to an embodiment of this application.

[0044] Figure 4 shows a schematic flowchart of step S103 in a method for predicting an anti-aging drug combination according to an embodiment of this application.

[0045] Figure 5 shows a block diagram of a predictive device for an anti-aging drug combination according to an embodiment of this application.

[0046] Figure 6 shows a schematic diagram of a computer device according to an embodiment of this application.

[0047] Figure 7 shows another structural schematic diagram of a computer device according to an embodiment of this application. Detailed Implementation Methods

[0048] To better understand the technical solution of this application, the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0049] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0050] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0051] To address the technical challenges of lacking models and methods that fully utilize multi-omics information and individual multi-omics expression profiles for prediction, this application proposes a prediction method, device, equipment, and medium for anti-aging drug combinations.

[0052] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0053] It should be noted that the embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0054] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0055] Please refer to Figure 1, which shows a schematic flowchart of a method for predicting anti-aging drug combinations according to an embodiment of this application. This method for predicting anti-aging drug combinations addresses the technical problem in related technologies, such as the lack of models and methods that fully utilize multi-omics information and individual multi-omics expression profiles for prediction.

[0056] As shown in Figure 1, the flow of a method for predicting an anti-aging drug combination according to an embodiment of this application includes:

[0057] Step S101: Obtain the individual multi-omics expression profile of the target drug user group, wherein the multi-omics includes genomics, RNAiomics, proteomics, post-translational modification genome, and metabolomics;

[0058] In this embodiment, multi-omics expression profile data of human bodies of different ages, i.e., different degrees of aging, can be collected, as well as individual multi-omics expression profiles of the target drug user group.

[0059] Step S102: Based on the individual multi-omics expression profile and the perturbation prediction model of the pre-trained multi-omics expression profile, predict the perturbation parameters of the target multi-omics corresponding to the target drug combination at different doses;

[0060] The perturbation prediction model based on pre-trained multi-omics expression profiles is used to predict the perturbation parameters of individual multi-omics for the target drug combination in the target drug user group. Different drug doses are set, and the corresponding perturbation parameters will also be different depending on the drug dose.

[0061] Step S103: Based on the perturbation parameters of the target multi-omics at each dose of the target drug combination, determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose.

[0062] In this embodiment, a pre-trained perturbation prediction model of multi-omics expression profiles is used to make predictions by making full use of multi-omics information and individual multi-omics expression profiles. This allows for the comprehensive, efficient, and accurate discovery of effective anti-aging drugs from a systems biology perspective. The drug combination can comprehensively regulate all pathways and targets in the human body, thereby improving the occurrence and development of age-related diseases.

[0063] As shown in Figure 2, in one embodiment, optionally, before step S101, the method further includes:

[0064] Step S201: Obtain the original multi-omics expression profiles of each cell line and the multi-omics perturbation information of various drugs at different doses on different human tissue cells;

[0065] In this embodiment, raw multi-omics (including genome, RNA, proteome, post-translational modification genome, and metabolome) expression profiles of cell lines and information on the changes in multi-omics expression levels of various drugs at different doses on different human tissue cells (i.e., the perturbation effect of drugs on targets) can be collected from public databases or experimental sources. Multi-omics expression profiles can reflect the molecular characteristics and functional state of cell lines, and perturbation information can indicate the biological pathways and mechanisms of drug effects. These data can provide useful features for predicting the synergistic effects of drug combinations in different cellular environments.

[0066] Step S202: Based on the original multi-omics expression profile and the multi-omics perturbation information, train the preset adversarial residual cell network model to obtain the perturbation prediction model of the multi-omics expression profile.

[0067] As shown in Figure 3, in one embodiment, optionally, step S202 includes:

[0068] Step S301: The full N-dimensional multi-omics original expression spectrum is encoded by an encoder to obtain the first n-dimensional latent vector used to characterize the expression level of the target.

[0069] Assuming that Xi represents the original expression levels of the entire N-dimensional genome, RNA, proteome, and metabolome, an encoder encodes Xi into an n-dimensional latent vector Zi (representing the expression level of the target).

[0070] Step S302: Characterize the cell lines corresponding to the original expression profiles of the multi-omics using Se ResNet, and generate a second n-dimensional hidden vector for characterizing the cell lines;

[0071] Cell lines derived from Xi are characterized using Se ResNet, and an n-dimensional hidden vector Zci is generated to represent the cell lines.

[0072] Step S303: Encode drug molecules using a graph neural network model to obtain drug features and generate a third n-dimensional hidden vector to characterize the drug.

[0073] In one embodiment, optionally, step S303 includes:

[0074] Drug molecules are encoded using a graph neural network model to obtain drug features, resulting in an m-dimensional hidden vector.

[0075] Drug dosage is represented by a one-dimensional vector, and drug dosage and drug features are concatenated into an m+1 dimensional latent vector.

[0076] The m+1 dimensional latent vector is transformed into the third n-dimensional latent vector using a multilayer perceptron.

[0077] In this embodiment, a graph neural network such as RdKit is used to encode drug molecules (including structural and physicochemical properties) to obtain drug features and generate an m-dimensional latent vector Gi. Then, a one-dimensional vector Si is used to represent the drug dosage, and the two vectors are combined to form an m+1-dimensional latent vector Zdi (representing the drug). A multilayer perceptron is then used to convert the m+1 dimension into an n-dimensional vector.

[0078] Step S304: Based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector, use the decoder to generate an N-dimensional total feature vector;

[0079] In one embodiment, optionally, step S304 includes:

[0080] Add the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector to obtain the total n-dimensional latent vector;

[0081] Based on the total n-dimensional latent vector, the decoder generates an N-dimensional total feature vector.

[0082] In this way, the potential information from drugs, multi-omics expression profiles, and cell lines is combined for the next step of model training.

[0083] Step S305: The N-dimensional total feature vector is used as input, and the corresponding multi-omics perturbation information is used as output vector to train the preset prediction model to obtain the perturbation prediction model of the multi-omics expression spectrum.

[0084] Specifically, different cell lines or different drugs can be divided into training sets and test sets to ensure that there are no duplicate cell lines or drugs between the training set and the test set. After multiple rounds of iteration to obtain a mature model, when a new drug and dosage and a new cell line are input, the perturbation effect of the drug dosage on the multi-omics of the cell line can be accurately predicted, such as changes in the expression levels of the genome, RNA, proteome, post-translational modification group, and metabolome.

[0085] As shown in Figure 4, in one embodiment, optionally, step S103 includes:

[0086] Step S401: Obtain the non-drug multi-omics expression profile of the young non-drug group corresponding to the target drug user group;

[0087] Step S402: Based on the perturbation parameters of the target multi-omics at each dose of the target drug combination, calculate the similarity score between the perturbation multi-omics expression profile of the target drug combination at each dose and the untreated multi-omics expression profile.

[0088] Step S403: Determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the similarity score.

[0089] In one embodiment, optionally, determining whether the target drug combination is a potential anti-aging drug combination and the corresponding target dosage based on the similarity score includes:

[0090] The similarity scores of the target drug combinations at various dosages are sorted in descending order.

[0091] Target drug combinations with similarity scores greater than preset scores and their corresponding dosages are identified as potential anti-aging drug combinations and target dosages.

[0092] In this embodiment, specifically, a similarity score can be calculated to measure the target drug user group. For example, the difference between the multi-omics expression profile of the aging group after drug perturbation and the multi-omics expression profile of the young group without drug perturbation. The smaller the difference, that is, the closer the multi-omics expression profile of the aging group after drug perturbation is to that of the young group, the higher the score of the drug or drug combination and its specific dosage. This is considered as screening out potential anti-aging drug combinations and dosages. In this way, by making full use of multi-omics information and individual multi-omics expression profiles for prediction, effective anti-aging drugs can be discovered comprehensively, efficiently, and accurately from a systems biology perspective. Drug combinations that comprehensively regulate all pathways and targets in the human body can improve the occurrence and development of age-related diseases.

[0093] Figure 5 shows a block diagram of a predictive device for an anti-aging drug combination according to an embodiment of this application.

[0094] As shown in Figure 5, in a second aspect, embodiments of this application provide a predictive device 50 for anti-aging drug combinations, comprising:

[0095] The first acquisition module 51 is used to acquire the individual multi-omics expression profile of the target drug user group, wherein the multi-omics includes genomics, RNAiome, proteomics, post-translational modification genome and metabolomics;

[0096] Prediction module 52 is used to predict the perturbation parameters of the target multi-omics corresponding to different doses of the target drug combination based on the individual multi-omics expression profile and the perturbation prediction model of the pre-trained multi-omics expression profile.

[0097] The determination module 53 is used to determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the perturbation parameters of the target multi-omics at each dose of the target drug combination.

[0098] In one embodiment, optionally, the apparatus further includes:

[0099] The second acquisition module is used to acquire the original multi-omics expression profiles of various cell lines and the multi-omics perturbation information of various drugs at different doses on different human tissue cells;

[0100] The training module is used to train a preset adversarial residual cell network model based on the original multi-omics expression profile and the multi-omics perturbation information to obtain a perturbation prediction model of the multi-omics expression profile.

[0101] In one embodiment, optionally, the training module includes:

[0102] The first coding unit is used to encode the full N-dimensional multi-omics original expression spectrum through the encoder to obtain the first n-dimensional latent vector used to characterize the expression level of the target.

[0103] A generation unit is used to characterize the cell lines corresponding to the original expression profiles of the multi-omics using SeResNet, and to generate a second n-dimensional hidden vector for characterizing the cell lines.

[0104] The second encoding unit is used to encode drug molecules through a graph neural network model to obtain drug features and generate a third n-dimensional hidden vector to characterize the drug.

[0105] The decoding unit is used to generate an N-dimensional total feature vector based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector using the decoder.

[0106] The model training unit is used to train the preset prediction model by taking the N-dimensional total feature vector as input and the corresponding multi-omics perturbation information as output vector, so as to obtain the perturbation prediction model of the multi-omics expression spectrum.

[0107] In one embodiment, optionally, the second encoding unit is used for:

[0108] Drug molecules are encoded using a graph neural network model to obtain drug features, resulting in an m-dimensional hidden vector.

[0109] Drug dosage is represented by a one-dimensional vector, and drug dosage and drug features are concatenated into an m+1 dimensional latent vector.

[0110] The m+1 dimensional latent vector is transformed into the third n-dimensional latent vector using a multilayer perceptron.

[0111] In one embodiment, optionally, the decoding unit is used for:

[0112] Add the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector to obtain the total n-dimensional latent vector;

[0113] Based on the total n-dimensional latent vector, the decoder generates an N-dimensional total feature vector.

[0114] In one embodiment, optionally, the determining module includes:

[0115] The acquisition unit is used to acquire the non-drug multi-omics expression profile of the young non-drug group corresponding to the target drug user group;

[0116] The calculation unit is used to calculate the similarity score between the perturbation parameters of the target multi-omics at each dose of the target drug combination and the untreated multi-omics expression profile, based on the perturbation parameters of the target multi-omics at each dose of the target drug combination.

[0117] The result determination unit determines whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the similarity score.

[0118] In one embodiment, optionally, the result determination unit is used for:

[0119] The similarity scores of the target drug combinations at various dosages are sorted in descending order.

[0120] Target drug combinations with similarity scores greater than preset scores and their corresponding dosages are identified as potential anti-aging drug combinations and target dosages.

[0121] Specific limitations regarding the predictive device for anti-aging drug combinations can be found in the limitations of the predictive method for anti-aging drug combinations described above, and will not be repeated here. Each module in the aforementioned predictive device for anti-aging drug combinations can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0122] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram is shown in Figure 6. The computer device includes a processor, memory, a network interface, and a database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a server-side method for predicting anti-aging drug combinations.

[0123] In one embodiment, a computer device is provided, which can be a client, and its internal structure diagram is shown in Figure 7. The computer device includes a processor, memory, network interface, display screen, and input device connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a client-side method for predicting anti-aging drug combinations.

[0124] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0125] The computer device in this application embodiment exists in various forms, including but not limited to:

[0126] (1) Mobile communication devices: These devices are characterized by their mobile communication capabilities and primarily aim to provide voice and data communication. These terminals include: smartphones (e.g., iPhones), multimedia phones, feature phones, and low-end phones, etc.

[0127] (2) Ultra-mobile personal computer devices: These devices fall under the category of personal computers, possessing computing and processing capabilities, and generally also have mobile internet access features. These terminals include PDAs, MIDs, and UMPCs, such as the iPad.

[0128] (3) Portable entertainment devices: These devices can display and play multimedia content. This category includes audio and video players (such as iPods), handheld game consoles, e-book readers, as well as smart toys and portable car navigation devices.

[0129] (4) Server: A device that provides computing services. The components of a server include a processor, hard disk, memory, system bus, etc. Servers are similar to general computer architectures, but because they need to provide highly reliable services, they have higher requirements in terms of processing power, stability, reliability, security, scalability, and manageability.

[0130] (5) Other electronic devices with data interaction functions.

[0131] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0132] Obtain individual multi-omics expression profiles of target drug users, wherein the multi-omics includes genomics, RNAiomics, proteomics, post-translational modification genomics, and metabolomics;

[0133] Based on the individual multi-omics expression profile and the perturbation prediction model of the pre-trained multi-omics expression profile, the perturbation parameters of the target multi-omics corresponding to the target drug combination at different doses are predicted.

[0134] Based on the perturbation parameters of the target multi-omics at various doses of the target drug combination, it is determined whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose.

[0135] In one embodiment, optionally, the method further includes:

[0136] To obtain the original multi-omics expression profiles of various cell lines and the multi-omics perturbation information of various drugs at different doses on different human tissue cells;

[0137] Based on the original multi-omics expression profile and the multi-omics perturbation information, a preset adversarial residual cell network model is trained to obtain a perturbation prediction model of the multi-omics expression profile.

[0138] In one embodiment, optionally, a preset prediction model is trained based on the original multi-omics expression profile and the multi-omics perturbation information to obtain a perturbation prediction model of the multi-omics expression profile, including:

[0139] The full N-dimensional multi-omics raw expression spectrum is encoded by an encoder to obtain the first n-dimensional latent vector used to characterize the expression level of the target.

[0140] The cell lines corresponding to the original multi-omics expression profiles are characterized by Se ResNet, and a second n-dimensional hidden vector is generated to characterize the cell lines.

[0141] Drug molecules are encoded using a graph neural network model to obtain drug features and generate a third n-dimensional hidden vector to characterize the drug.

[0142] Based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector, the decoder generates an N-dimensional total feature vector.

[0143] The N-dimensional total feature vector is used as input, and the corresponding multi-omics perturbation information is used as output vector to train the preset prediction model, so as to obtain the perturbation prediction model of the multi-omics expression spectrum.

[0144] In one embodiment, optionally, drug molecules are encoded using a graph neural network model to obtain drug features and generate a third n-dimensional latent vector for characterizing the drug, including:

[0145] Drug molecules are encoded using a graph neural network model to obtain drug features, resulting in an m-dimensional hidden vector.

[0146] Drug dosage is represented by a one-dimensional vector, and drug dosage and drug features are concatenated into an m+1 dimensional latent vector.

[0147] The m+1 dimensional latent vector is transformed into the third n-dimensional latent vector using a multilayer perceptron.

[0148] In one embodiment, optionally, an N-dimensional total feature vector is generated using a decoder based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector, including:

[0149] Add the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector to obtain the total n-dimensional latent vector;

[0150] Based on the total n-dimensional latent vector, the decoder generates an N-dimensional total feature vector.

[0151] In one embodiment, optionally, determining whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the perturbation parameters of the target multi-omics at various doses of the target drug combination includes:

[0152] Obtain the non-drug multi-omics expression profile of the young non-drug group corresponding to the target drug user group;

[0153] Based on the perturbation parameters of the target multi-omics at each dose of the target drug combination, the similarity score between the perturbation multi-omics expression profile of the target drug combination at each dose and the untreated multi-omics expression profile is calculated.

[0154] The similarity score is used to determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose.

[0155] In one embodiment, optionally, determining whether the target drug combination is a potential anti-aging drug combination and the corresponding target dosage based on the similarity score includes:

[0156] The similarity scores of the target drug combinations at various dosages are sorted in descending order.

[0157] Target drug combinations with similarity scores greater than preset scores and their corresponding dosages are identified as potential anti-aging drug combinations and target dosages.

[0158] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or electronic device described above can be referred to the relevant descriptions in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0159] The technical solution of this application has been described in detail above with reference to the accompanying drawings. By using a pre-trained perturbation prediction model of multi-omics expression profiles, and making full use of multi-omics information and individual multi-omics expression profiles for prediction, effective anti-aging drugs can be discovered comprehensively, efficiently and accurately from the perspective of systems biology. Drug combinations that comprehensively regulate all pathways and targets in the human body can improve the occurrence and development of age-related diseases.

[0160] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0161] It should be understood that although the terms "first," "second," etc., may be used to describe the setting units in the embodiments of this application, these setting units should not be limited to these terms. These terms are only used to distinguish the setting units from each other. For example, without departing from the scope of the embodiments of this application, the first setting unit may also be referred to as the second setting unit, and similarly, the second setting unit may also be referred to as the first setting unit.

[0162] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0163] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0164] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional units.

[0165] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0166] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for predicting anti-aging drug combinations, characterized in that, include: The process involves obtaining individual multi-omics expression profiles of the target drug user group, where the multi-omics includes genomics, RNAi, proteomics, post-translational modification, and metabolomics; predicting perturbation parameters of the target multi-omics at different dosages based on the individual multi-omics expression profiles and a pre-trained perturbation prediction model of the multi-omics expression profiles; and determining whether the target drug combination is a potential anti-aging drug combination and its corresponding target dosage based on the perturbation parameters of the target multi-omics at each dosage of the target drug combination. The training process of the perturbation prediction model of the multi-omics expression spectrum includes: encoding the full N-dimensional original multi-omics expression spectrum through an encoder to obtain the first n-dimensional latent vector used to characterize the expression level of the target. Cell lines corresponding to the original multi-omics expression profiles are characterized using SeResNet, generating a second n-dimensional hidden vector to characterize the cell lines. Drug molecules are encoded using a graph neural network model to obtain drug features, generating a third n-dimensional hidden vector to characterize the drugs. Based on the first, second, and third n-dimensional hidden vectors, an N-dimensional total feature vector is generated using a decoder. The N-dimensional total feature vector is used as input, and the corresponding multi-omics perturbation information is used as the output vector to train a preset prediction model to obtain a perturbation prediction model for the multi-omics expression profiles.

2. The method for predicting anti-aging drug combinations according to claim 1, characterized in that, The method further includes: acquiring the original multi-omics expression profiles of each cell line and the multi-omics perturbation information of various drugs at different doses on different human tissue cells; and training a preset adversarial residual cell network model based on the original multi-omics expression profiles and the multi-omics perturbation information to obtain a perturbation prediction model of the multi-omics expression profiles.

3. The method for predicting anti-aging drug combinations according to claim 1, characterized in that, Encoding drug molecules using a graph neural network model to obtain drug features and generating a third n-dimensional latent vector to characterize the drug includes: encoding drug molecules using a graph neural network model to obtain drug features, resulting in an m-dimensional latent vector; representing drug dosage with a one-dimensional vector and concatenating drug dosage and drug features into an m+1-dimensional latent vector; and converting the m+1-dimensional latent vector into the third n-dimensional latent vector using a multilayer perceptron.

4. The method for predicting anti-aging drug combinations according to claim 1, characterized in that, Based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector, an N-dimensional total feature vector is generated using a decoder, including: adding the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector to obtain a total n-dimensional latent vector; and generating an N-dimensional total feature vector using the decoder based on the total n-dimensional latent vector.

5. The method for predicting anti-aging drug combinations according to claim 1, characterized in that, Determining whether the target drug combination is a potential anti-aging drug combination and its corresponding target dose based on the perturbation parameters of the target multi-omics at various doses of the target drug combination includes: obtaining the untreated multi-omics expression profile of the young untreated group corresponding to the target drug treatment group; calculating the similarity score between the perturbation multi-omics expression profile of the target drug combination at each dose and the untreated multi-omics expression profile based on the perturbation parameters of the target multi-omics at various doses of the target drug combination; and determining whether the target drug combination is a potential anti-aging drug combination and its corresponding target dose based on the similarity score.

6. The method for predicting anti-aging drug combinations according to claim 5, characterized in that, Determining whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the similarity score includes: sorting the similarity scores of the target drug combination at each dose in descending order; and determining the target drug combination and the corresponding dose with a similarity score greater than a preset score as a potential anti-aging drug combination and target dose.

7. A predictive device for anti-aging drug combinations, characterized in that, include: The first acquisition module is used to acquire the individual multi-omics expression profile of the target drug user group, wherein the multi-omics includes genomics, RNAiome, proteomics, post-translational modification genome and metabolomics; The prediction module is used to predict the perturbation parameters of the target multi-omics corresponding to different doses of the target drug combination based on the individual multi-omics expression profile and the perturbation prediction model of the pre-trained multi-omics expression profile. A determination module is used to determine whether the target drug combination is a potential anti-aging drug combination and the corresponding target dose based on the perturbation parameters of the target multi-omics at various doses of the target drug combination. The training process of the perturbation prediction model for the multi-omics expression profile includes: encoding the full N-dimensional original multi-omics expression profile using an encoder to obtain a first n-dimensional latent vector characterizing the expression level of the target; characterizing the cell line corresponding to the original multi-omics expression profile using SeResNet to generate a second n-dimensional latent vector characterizing the cell line; encoding drug molecules using a graph neural network model to obtain drug features and generating a third n-dimensional latent vector characterizing the drug; generating an N-dimensional total feature vector using a decoder based on the first n-dimensional latent vector, the second n-dimensional latent vector, and the third n-dimensional latent vector; and training a preset prediction model using the N-dimensional total feature vector as input and the corresponding multi-omics perturbation information as output vector to obtain the perturbation prediction model for the multi-omics expression profile.

8. A computer device, characterized in that, include: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being configured to perform the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The device stores computer-executable instructions for performing the method as described in any one of claims 1 to 6.

Images