Cancer prognosis evaluation and target recognition method and system based on deep neural network

Abstract

The invention discloses a cancer prognosis evaluation and target recognition method based on a deep neural network, and the method comprises the steps: obtaining gene expression data, and carrying out the preprocessing of the gene expression data, and obtaining initial data; performing feature extraction processing on the initial data by using a deep neural network combining a diffusion model and an auto-encoder to obtain low-dimensional representation of gene expression features; inputting the low-dimensional expression of the gene expression characteristics into an XGBoost survival analysis model to obtain a risk value, and obtaining a cancer prognosis evaluation result according to the risk value; and performing differential expression analysis according to the risk value and the gene expression of the patient to obtain the cancer prognosis target. Visibly, according to the cancer prognosis risk assessment method and system, through the deep neural network combining the diffusion model and the auto-encoder and the XGBoost survival analysis model, accurate assessment of the prognosis risk of the cancer patient and recognition of the related target spot are achieved, and therefore a scientific basis is provided for design of personalized treatment and targeted therapy.

Description

Technical Field

[0001] The present invention relates to the technical field of cancer prognosis analysis, and in particular to a method and system for cancer prognosis assessment and target identification based on a deep neural network. Background Art

[0002] Cancer is a complex and highly lethal malignant disease whose pathogenesis is affected by the interaction between genomic characteristics and environmental factors. Although some progress has been made in the research and treatment of cancer, the prognosis of different patients still varies significantly due to the biological characteristics of individual tumors. Therefore, accurately assessing the prognostic risk of cancer remains a major challenge in clinical practice. In recent years, through advanced deep learning technology, important information can be extracted from relevant biological data to improve the accuracy of prognostic assessment. At the same time, accurate assessment of cancer prognostic risk helps to accurately discover potential targets related to it, and provide assistance for the implementation of personalized treatment and the design of targeted therapies.

[0003] However, it is currently very difficult to extract key tumor feature information from complex gene expression data, and in the biomedical field, insufficient data or scarce samples often occur, and obtaining sufficient data is often a challenge. It can be seen that the existing technology has defects that need to be solved urgently. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a cancer prognosis assessment and target identification method and system based on deep neural network, which realizes accurate assessment of the prognosis risk of cancer patients and identification of relevant targets by combining the deep neural network of diffusion model and autoencoder and XGBoost survival analysis model, thereby providing a scientific basis for the design of personalized treatment and targeted therapy.

[0005] In order to solve the above technical problems, the first aspect of the present invention discloses a cancer prognosis assessment and target identification method based on a deep neural network, the method comprising:

[0006] A cancer prognosis assessment and target identification method based on deep neural network, characterized in that the method comprises:

[0007] Acquiring gene expression data, and preprocessing the gene expression data to obtain initial data;

[0008] Performing feature extraction processing on the initial data using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional expression of gene expression features;

[0009] Inputting the low-dimensional expression of the gene expression feature into the XGBoost survival analysis model to obtain a risk value, and obtaining a cancer prognosis evaluation result according to the risk value;

[0010] Differential expression analysis is performed based on the risk value and the patient's gene expression to obtain cancer prognosis targets.

[0011] As an optional embodiment, in the first aspect of the present invention, the obtaining of gene expression data and preprocessing of the gene expression data to obtain initial data includes: obtaining the gene expression data of the patient and filling the missing values ​​in the gene expression data with the average value to obtain initial data.

[0012] As an optional embodiment, in the first aspect of the present invention, the use of a deep neural network combining a diffusion model and an autoencoder to perform feature extraction processing on the initial data to obtain a low-dimensional expression of gene expression features includes:

[0013] The initial data is mapped to a low-dimensional space, key features are extracted, and a low-dimensional expression of gene expression features is obtained.

[0014] As an optional implementation, in the first aspect of the present invention, the training process of the deep neural network combining the diffusion model and the autoencoder comprises the following steps:

[0015] Acquiring experimental data, and preprocessing the experimental data to obtain training data;

[0016] adding noise to the training data to implement a diffusion step, returning the noisy data and the generated noise to obtain sample data;

[0017] Inputting the sample data into a deep neural network combining a diffusion model and an autoencoder, wherein the deep neural network combining a diffusion model and an autoencoder comprises an encoder and a decoder, wherein the encoder maps the sample data to a latent space, and the decoder restores the compressed representation to the original dimension to obtain a low-dimensional expression of the training gene expression feature;

[0018] Use the optimizer to perform backpropagation, calculate the gradient of the loss function on the parameters of the deep neural network that combines the diffusion model and the autoencoder, minimize the reconstruction error, and update the parameters of the encoder and the decoder;

[0019] The deep neural network combining the diffusion model and the autoencoder is trained according to the number of cycles, and each batch performs the diffusion step, forward propagation, loss calculation, backpropagation and optimization.

[0020] As an optional implementation, in the first aspect of the present invention, the optimizer is an Adam optimizer.

[0021] As an optional implementation, in the first aspect of the present invention, the step of obtaining experimental data and preprocessing the experimental data to obtain training data includes:

[0022] Gene expression data and survival data of cancer patients were obtained from TCGA and GEO public data sets, and the missing values ​​in the obtained data were filled with the average to obtain training data.

[0023] As an optional implementation, in the first aspect of the present invention, the training process of the XGBoost survival analysis model includes the following steps:

[0024] The low-dimensional expression of the training gene expression feature is used as input to construct a data format suitable for survival analysis;

[0025] Initialize an XGBoost survival analysis model, and use the low-dimensional expression and survival time of the training gene expression features as labels to train the XGBoost survival analysis model.

[0026] The second aspect of the present invention discloses a cancer prognosis assessment and target identification system based on a deep neural network, the system comprising:

[0027] A data acquisition module, the data acquisition module is used to acquire gene expression data and pre-process the gene expression data to obtain initial data;

[0028] A feature acquisition module, wherein the feature acquisition module is used to perform feature extraction processing on the initial data using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional expression of gene expression features;

[0029] An evaluation result acquisition module, wherein the evaluation result acquisition module is used to input the low-dimensional expression of the gene expression feature into the XGBoost survival analysis model to obtain a risk value, and obtain a cancer prognosis evaluation result according to the risk value;

[0030] A target acquisition module is used to perform differential expression analysis based on the risk value and the patient's gene expression to obtain cancer prognosis targets.

[0031] As an optional embodiment, in the second aspect of the present invention, the data acquisition module acquires gene expression data and preprocesses the gene expression data to obtain initial data, including: acquiring the gene expression data of the patient and filling the missing values ​​in the gene expression data with the average value to obtain initial data.

[0032] As an optional implementation, in the second aspect of the present invention, the feature acquisition module uses a deep neural network combining a diffusion model and an autoencoder to perform feature extraction processing on the initial data to obtain a low-dimensional expression of gene expression features, including:

[0033] The initial data is mapped to a low-dimensional space, key features are extracted, and a compressed representation of gene expression features is obtained.

[0034] As an optional implementation, in the second aspect of the present invention, the training process of the deep neural network combining the diffusion model and the autoencoder comprises the following steps:

[0035] Acquiring experimental data, and preprocessing the experimental data to obtain training data;

[0036] adding noise to the training data to implement a diffusion step, returning the noisy data and the generated noise to obtain sample data;

[0037] Inputting the sample data into a deep neural network combining a diffusion model and an autoencoder, wherein the deep neural network combining a diffusion model and an autoencoder comprises an encoder and a decoder, wherein the encoder maps the sample data to a latent space, and the decoder restores the compressed representation to the original dimension to obtain a low-dimensional expression of the training gene expression feature;

[0038] Use the optimizer to perform backpropagation, calculate the gradient of the loss function on the parameters of the deep neural network that combines the diffusion model and the autoencoder, minimize the reconstruction error, and update the parameters of the encoder and the decoder;

[0039] The deep neural network combining the diffusion model and the autoencoder is trained according to the number of cycles, and each batch performs the diffusion step, forward propagation, loss calculation, backpropagation and optimization.

[0040] As an optional implementation, in the second aspect of the present invention, the optimizer is an Adam optimizer.

[0041] As an optional implementation, in the second aspect of the present invention, the step of acquiring experimental data and preprocessing the experimental data to obtain training data includes:

[0042] Gene expression data and survival data of cancer patients were obtained from TCGA and GEO public data sets, and the missing values ​​in the obtained data were filled with the average to obtain training data.

[0043] As an optional implementation, in the second aspect of the present invention, the training process of the XGBoost survival analysis model includes the following steps:

[0044] The low-dimensional expression of the training gene expression feature is used as input to construct a data format suitable for survival analysis;

[0045] Initialize an XGBoost survival analysis model, and use the low-dimensional expression and survival time of the training gene expression features as labels to train the XGBoost survival analysis model.

[0046] The third aspect of the present invention discloses another cancer prognosis assessment and target identification device based on a deep neural network, the device comprising:

[0047] A memory storing executable program code;

[0048] a processor coupled to the memory;

[0049] The processor calls the executable program code stored in the memory to execute part or all of the steps in a cancer prognosis assessment and target identification method based on a deep neural network disclosed in the first aspect of an embodiment of the present invention.

[0050] A fourth aspect of an embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute part or all of the steps in a cancer prognosis assessment and target identification method based on a deep neural network disclosed in the first aspect of an embodiment of the present invention.

[0051] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0052] In an embodiment of the present invention, gene expression data is obtained, and the gene expression data is preprocessed to obtain initial data; the initial data is subjected to feature extraction processing using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional representation of gene expression features; the gene expression features are input into an XGBoost survival analysis model to obtain a cancer prognosis evaluation result; and gene differential expression analysis is performed based on the cancer prognosis evaluation result to obtain cancer prognosis targets. It can be seen that the present invention can effectively extract key tumor feature information from complex gene expression data through the combination of a diffusion model and an autoencoder, thereby generating high-quality synthetic data. On this basis, by introducing the XGBoost model, the accuracy of cancer prognosis evaluation can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0054] Figure 1 It is a flowchart of a method for cancer prognosis assessment and target identification based on a deep neural network disclosed in an embodiment of the present invention;

[0055] Figure 2 It is a schematic diagram of the structure of a cancer prognosis assessment and target identification system based on a deep neural network disclosed in an embodiment of the present invention;

[0056] Figure 3 It is a schematic structural diagram of another cancer prognosis assessment and target identification device based on a deep neural network disclosed in an embodiment of the present invention;

[0057] Figure 4 This is an analysis chart of the differential expression results of skin cancer prognostic targets. DETAILED DESCRIPTION

[0058] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0060] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0061] The present invention discloses a cancer prognosis assessment and target identification method and system based on a deep neural network, which can effectively extract key tumor feature information from complex gene expression data by combining a diffusion model with an autoencoder, thereby generating high-quality synthetic data. On this basis, the XGBoost model is introduced to effectively improve the accuracy of cancer prognosis assessment. The following are detailed descriptions.

[0062] Embodiment 1

[0063] See also Figure 1 , Figure 1 : is a flow chart of a method for cancer prognosis assessment and target identification based on a deep neural network disclosed in an embodiment of the present invention. Figure 1 The described method is applied to a cancer prognosis assessment and target recognition system based on a deep neural network. The recognition system can be a corresponding recognition terminal, recognition device or server, and the server can be a local server or a cloud server, which is not limited in the embodiments of the present invention. Figure 1 As shown, the cancer prognosis assessment and target identification method based on deep neural network may include the following operations:

[0064] 101. Obtain gene expression data, and preprocess the gene expression data to obtain initial data.

[0065] 102. Use a deep neural network that combines a diffusion model and an autoencoder to perform feature extraction processing on the initial data to obtain a low-dimensional expression of gene expression features.

[0066] 103. Input the low-dimensional expression of the gene expression feature into the XGBoost survival analysis model to obtain a risk value, and obtain a cancer prognosis evaluation result based on the risk value; wherein, it is divided into a high-risk group and a low-risk group according to the median of the risk value.

[0067] 104. Perform differential expression analysis based on the risk value and the patient's gene expression to obtain cancer prognosis targets.

[0068] It can be seen that the method described in the embodiment of the present invention can achieve accurate assessment of the prognosis risk of cancer patients and identification of relevant targets through the deep neural network combining the diffusion model and the autoencoder and the XGBoost survival analysis model, thereby providing a scientific basis for the design of personalized treatment and targeted therapy. And by using the deep learning network combining the diffusion model and the autoencoder to effectively extract key tumor feature information from complex gene expression data, it is possible to generate high-quality synthetic data. On this basis, by introducing the XGBoost model, the accuracy of cancer prognosis assessment can be effectively improved.

[0069] In an optional embodiment, the step of obtaining gene expression data and preprocessing the gene expression data to obtain initial data in step 101 includes: obtaining gene expression data of a patient and filling missing values ​​in the gene expression data with average values ​​to obtain initial data.

[0070] In an optional embodiment, the step 102 of performing feature extraction processing on the initial data using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional expression of gene expression features includes:

[0071] The initial data is mapped to a low-dimensional space, key features are extracted, and a low-dimensional expression of gene expression features is obtained.

[0072] In the embodiment of the present invention, the deep neural network combining the diffusion model and the autoencoder can reconstruct the gene expression data of cancer cells, obtain higher quality data, and more effectively extract discriminative features for risk assessment of cancer prognosis.

[0073] In an optional embodiment, the training process of the deep neural network combining the diffusion model and the autoencoder in step 102 includes the following steps:

[0074] Acquiring experimental data, and preprocessing the experimental data to obtain training data;

[0075] adding noise to the training data to implement a diffusion step, returning the noisy data and the generated noise to obtain sample data;

[0076] Inputting the sample data into a deep neural network combining a diffusion model and an autoencoder, wherein the deep neural network combining a diffusion model and an autoencoder comprises an encoder and a decoder, wherein the encoder maps the sample data to a latent space, and the decoder restores the compressed representation to the original dimension to obtain a low-dimensional expression of the training gene expression feature;

[0077] Use the Adam optimizer for back propagation, calculate the gradient of the loss function with respect to the parameters of the deep neural network that combines the diffusion model and the autoencoder, minimize the reconstruction error, and update the parameters of the encoder and the decoder;

[0078] The deep neural network combining the diffusion model and the autoencoder is trained according to the number of cycles, and each batch performs the diffusion step, forward propagation, loss calculation, backpropagation and optimization.

[0079] In an embodiment of the present invention, the training data is input into a deep neural network that combines a diffusion model and an autoencoder, wherein a customized autoencoder structure is defined, including an encoder and a decoder; the diffusion step is implemented by noise, and the noisy data and the generated noise are returned. Then the input dimension (i.e., the number of genes) is obtained, and an autoencoder model is created. The encoder maps the training data x to the latent space, the encoder E outputs the latent variable z, and the decoder D restores the compressed representation back to the original dimension to obtain the reconstruction result x′. The mean square error loss and the Adam optimizer are defined to calculate the loss function between the reconstructed output and the original output, the denoising loss, and other loss functions. Back propagation calculates the gradient of the loss function to the model parameters, minimizes the loss function, and updates the parameters of the encoder and the decoder. The model is trained according to the number of cycles, and each batch performs the diffusion step, forward propagation, loss calculation, back propagation, and optimization. Among them, the autoencoder loss is the sum of the reconstruction loss, denoising loss, and regularization loss, and the expression of the reconstruction loss is as follows:

[0080]

[0081] The reconstruction loss is the mean square error, where x i is the i-th sample of the input data, is the i-th sample reconstructed by the decoder and N is the total number of samples.

[0082] The expression of denoising loss is as follows:

[0083]

[0084] The denoising loss is the mean square error used for denoising, where z i is the true noise-free latent representation, is the denoised result of the model prediction.

[0085] The regularization loss is L2 regularization loss, and its expression is as follows:

[0086]

[0087] Where W i is the i-th parameter of the model and λ is a hyperparameter of the regularization strength.

[0088] Therefore, the loss of the autoencoder is L total =αL recon +βL denoise +γL reg , where α, β, and γ are weight parameters used to balance the impact of each part of the loss.

[0089] In an optional embodiment, in the above steps, the step of obtaining experimental data and preprocessing the experimental data to obtain training data includes:

[0090] Gene expression data and survival data of cancer patients were obtained from TCGA and GEO public data sets, and the missing values ​​in the obtained data were filled with the average to obtain training data.

[0091] In an optional embodiment, the training process of the XGBoost survival analysis model in step 103 includes the following steps:

[0092] The low-dimensional expression of the training gene expression feature is used as input to construct a data format suitable for survival analysis;

[0093] Initialize an XGBoost survival analysis model, and use the low-dimensional expression and survival time of the training gene expression features as labels to train the XGBoost survival analysis model.

[0094] In the embodiment of the present invention, the low-dimensional expression of the training gene expression features reconstructed by the deep neural network combining the diffusion model and the autoencoder is input to construct a data format suitable for survival analysis, initialize the model and use the feature data and survival time to train the model. The goal is to minimize the sum of the loss function and the difficulty of setting the regularization term, that is, to minimize the objective function, so as to continuously optimize the model. Among them, the objective function of the XBboost survival analysis model can be expressed as Among them, l is the loss function, y′ i is the predicted value, y i is the true value; the setting of the regular term is related to the complexity of the model and can be expressed as It contains two important parts. One is the number of leaf nodes T in the tree, and the other is the score of the leaf nodes in the tree. (L2 regularization of w is equivalent to adding L2 smoothing to the score of each leaf node.) Therefore, the objective function of XGBoost is the loss function revealing the training error plus the complexity of the regularization definition, where the goal is to minimize the objective function L(Φ).

[0095] Based on the low-dimensional expression of the training gene expression features, the loss function and regularization definition complexity are calculated based on the XGBoost ensemble learning algorithm. The performance is continuously optimized by minimizing their sum (i.e., the objective function L(Φ)), thereby predicting the survival risk of each sample in the input data.

[0096] Finally, the accuracy of the cancer prognosis prediction model constructed by our algorithm was evaluated by calculating the C-index.

[0097] In this step, the low-dimensional expression of the training gene expression features is input into the commonly used Cox proportional hazard model, RSF random survival forest, Cox-elastic ne and DNN deep neural network for calculation and evaluation to compare the accuracy of the methods. The specific results are shown in Table 1. The confidence index obtained by our method is higher than that of the compared methods, which proves the superiority of our method.

[0098] Table 1 Confidence C-index scores of cancer public datasets

[0099]

[0100] Regarding differential expression analysis, the results of differential expression analysis in skin cancer are as follows: Figure 4 As shown. Differential expression analysis is an analytical method that identifies genes that change significantly in a certain state by comparing the differences in gene expression levels in different sample groups (such as cancer tissues and normal tissues). Differential expression analysis is of great significance in cancer target identification. It can not only help discover new cancer biomarkers for early diagnosis, prognosis evaluation and treatment effect monitoring of cancer, but also reveal potential drug targets by identifying significantly upregulated or downregulated genes, especially those key genes and pathways that play an important role in tumor occurrence and progression. In addition, differential expression analysis can deepen the understanding of cancer-related molecular mechanisms, help researchers identify the core factors that drive malignant transformation, and provide a strong scientific basis for targeted therapy. Therefore, this technology has irreplaceable application value in cancer research and therapeutic target development. The top 10 potential targets that can be predicted to affect the prognosis of skin cancer are shown in Table 2.

[0101] Table 2 Top 10 potential targets affecting skin cancer prognosis

[0102]

[0103]

[0104] The training data is input into the deep neural network combined with diffusion autoencoder, and the deep neural network combined with diffusion autoencoder is trained to output features. The XGBoost survival analysis model is used for training based on the features, and decision trees are built in multiple iterations. Each tree is improved based on the prediction errors of all previous trees. The model settings of the XBboost survival analysis model are determined, and the cancer prognosis is evaluated based on the genetic data.

[0105] Deep neural networks based on combining diffusion models and autoencoders are a strategy that combines the generative capabilities of diffusion models with the efficient encoding capabilities of autoencoders. Traditional autoencoders consist of an encoder and a decoder, where the encoder compresses the input data into a latent representation and the decoder reconstructs the original data from the latent representation. The diffusion model is a generative model based on a Markov chain that maps data to a simple prior distribution (such as a Gaussian distribution) by gradually adding noise, and then generates data from the noise through the inverse process.

[0106] Combining the advantages of diffusion models and autoencoders can significantly improve the ability to learn and generate data features. Diffusion models learn the underlying distribution of data by gradually adding noise. This method performs well in generating high-quality new samples, but is extremely sensitive to noise management and parameter settings. Autoencoders focus on effectively compressing data and extracting key features. Although they can learn quickly, they may ignore subtle structural information in the data in some cases, resulting in reduced reconstruction quality. Combining these two to form a diffusion autoencoder, the model can not only improve the robustness to noise through the diffusion step, but also improve the reconstruction quality, thereby extracting stronger feature representations. This combination is particularly important in the application of cancer prognosis assessment and target identification. First, the diffusion autoencoder can effectively extract key tumor feature information from complex gene expression data, thereby generating high-quality synthetic data, which is of great significance for situations where data is insufficient or samples are scarce, especially in the biomedical field, where obtaining sufficient data is often a challenge. Finally, the feature representation after dimensionality reduction helps to improve the prediction accuracy of the prognostic model, making it more practical and reliable, thereby providing support for identifying potential therapeutic targets. The comprehensive application of this method can significantly promote the progress of cancer research and improve treatment effects.

[0107] At the same time, the application of XGBoost in cancer prognosis assessment and target identification has shown many significant advantages. As an advanced ensemble learning method, XGBoost has strong predictive capabilities and can effectively handle high-dimensional features and complex data patterns, thereby providing highly accurate results in survival analysis and risk assessment. Feature selection is automatically performed during the training process. Based on the importance score of the features, XGBoost can identify key biomarkers that are closely related to the occurrence and development of cancer, which provides a solid foundation for target identification. On the other hand, XGBoost uses parallel computing and optimization algorithms, which enables it to show superior computing speed when processing large-scale clinical data, suitable for rapid response medical environments. XGBoost also effectively prevents overfitting and enhances the generalization ability of the model by introducing L1 and L2 regularization, which is especially important when processing biological data, because biological data is usually complex and susceptible to noise. Therefore, the application of XGBoost not only improves the accuracy of cancer prognosis assessment and helps doctors develop personalized treatment plans, but also effectively identifies key genes and pathways related to cancer, providing potential targets for new drug development and treatment strategies, thereby promoting the in-depth development of cancer research.

[0108] Embodiment 2

[0109] See also Figure 2 , Figure 2 is a schematic diagram of the structure of a cancer prognosis assessment and target identification system based on a deep neural network disclosed in an embodiment of the present invention. Figure 2 The described system can be applied to a corresponding identification terminal, identification device or server, and the server can be a local server or a cloud server, which is not limited in the embodiment of the present invention. Figure 2 As shown, the system may include:

[0110] A data acquisition module 100, the data acquisition module 100 is used to acquire gene expression data and pre-process the gene expression data to obtain initial data;

[0111] A feature acquisition module 200, wherein the feature acquisition module 200 is used to perform feature extraction processing on the initial data using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional expression of gene expression features;

[0112] An evaluation result acquisition module 300, wherein the evaluation result acquisition module 300 is used to input the gene expression feature into the XGBoost survival analysis model to obtain a risk value, and obtain a cancer prognosis evaluation result according to the risk value;

[0113] The target acquisition module 400 is used to perform differential expression analysis based on the risk value and the patient's gene expression to obtain cancer prognosis targets.

[0114] As an optional embodiment, in the second aspect of the present invention, the data acquisition module 100 acquires gene expression data and preprocesses the gene expression data to obtain initial data, including: acquiring the gene expression data of the patient and filling the missing values ​​in the gene expression data with the average value to obtain initial data.

[0115] As an optional implementation, in the second aspect of the present invention, the feature acquisition module 200 uses a deep neural network combining a diffusion model and an autoencoder to perform feature extraction processing on the initial data to obtain a low-dimensional expression of gene expression features, including:

[0116] The initial data is mapped to a low-dimensional space, key features are extracted, and gene expression features are obtained.

[0117] As an optional implementation, in the second aspect of the present invention, the training process of the deep neural network combining the diffusion model and the autoencoder comprises the following steps:

[0118] Acquiring experimental data, and preprocessing the experimental data to obtain training data;

[0119] adding noise to the training data to implement a diffusion step, returning the noisy data and the generated noise to obtain sample data;

[0120] Inputting the sample data into a deep neural network combining a diffusion model and an autoencoder, wherein the deep neural network combining a diffusion model and an autoencoder comprises an encoder and a decoder, wherein the encoder maps the sample data to a latent space, and the decoder restores the compressed representation to the original dimension to obtain a low-dimensional expression of the training gene expression feature;

[0121] Use the optimizer to perform backpropagation, calculate the gradient of the loss function on the parameters of the deep neural network that combines the diffusion model and the autoencoder, minimize the reconstruction error, and update the parameters of the encoder and the decoder;

[0122] The deep neural network combining the diffusion model and the autoencoder is trained according to the number of cycles, and each batch performs the diffusion step, forward propagation, loss calculation, backpropagation and optimization.

[0123] As an optional implementation, in the second aspect of the present invention, the optimizer is an Adam optimizer.

[0124] As an optional implementation, in the second aspect of the present invention, the step of acquiring experimental data and preprocessing the experimental data to obtain training data includes:

[0125] Gene expression data and survival data of cancer patients were obtained from TCGA and GEO public data sets, and the missing values ​​in the obtained data were filled with the average to obtain training data.

[0126] As an optional implementation, in the second aspect of the present invention, the training process of the XGBoost survival analysis model includes the following steps:

[0127] The low-dimensional expression of the training gene expression feature is used as input to construct a data format suitable for survival analysis;

[0128] Initialize an XGBoost survival analysis model, and use the low-dimensional expression and survival time of the training gene expression features as labels to train the XGBoost survival analysis model.

[0129] Embodiment 3

[0130] See also Figure 3 , Figure 3 is a schematic diagram of another cancer prognosis assessment and target identification system based on a deep neural network disclosed in an embodiment of the present invention. Figure 3 As shown, the device may include:

[0131] A memory 301 storing executable program codes;

[0132] a processor 302 coupled to the memory 301;

[0133] The processor 302 calls the executable program code stored in the memory 301 to execute part or all of the steps in the cancer prognosis assessment and target identification method based on deep neural network disclosed in the first embodiment of the present invention.

[0134] Embodiment 4

[0135] An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute some or all of the steps in the cancer prognosis assessment and target identification method based on deep neural network disclosed in Example 1 of the present invention.

[0136] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0137] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0138] Finally, it should be noted that the cancer prognosis assessment and target identification method and system based on deep neural network disclosed in the embodiment of the present invention discloses only the preferred embodiment of the present invention, which is only used to illustrate the technical scheme of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical schemes described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical schemes from the spirit and scope of the technical schemes of the embodiments of the present invention.

Claims

1. A cancer prognosis assessment and target identification method based on deep neural network, characterized in that: The method comprises: Acquiring gene expression data, and preprocessing the gene expression data to obtain initial data; Performing feature extraction processing on the initial data using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional expression of gene expression features; Inputting the low-dimensional expression of the gene expression feature into the XGBoost survival analysis model to obtain a risk value, and obtaining a cancer prognosis evaluation result according to the risk value; Differential expression analysis is performed based on the risk value and the patient's gene expression to obtain cancer prognosis targets.

2. The cancer prognosis assessment and target identification method based on deep neural network according to claim 1, characterized in that: The obtaining of gene expression data and preprocessing of the gene expression data to obtain initial data includes: obtaining gene expression data of a patient and filling missing values ​​in the gene expression data with average values ​​to obtain initial data.

3. The cancer prognosis assessment and target identification method based on deep neural network according to claim 1, characterized in that: The method of using a deep neural network combining a diffusion model and an autoencoder to perform feature extraction processing on the initial data to obtain a low-dimensional expression of gene expression features includes: The initial data is mapped to a low-dimensional space, key features are extracted, and a low-dimensional expression of gene expression features is obtained.

4. The method for cancer prognosis assessment and target identification based on deep neural network according to claim 1, characterized in that: The training process of the deep neural network combining the diffusion model and the autoencoder includes the following steps: Acquiring experimental data, and preprocessing the experimental data to obtain training data; adding noise to the training data to implement a diffusion step, returning the noisy data and the generated noise to obtain sample data; Inputting the sample data into a deep neural network combining a diffusion model and an autoencoder, wherein the deep neural network combining a diffusion model and an autoencoder comprises an encoder and a decoder, wherein the encoder maps the sample data to a latent space, and the decoder restores the compressed representation to the original dimension to obtain a low-dimensional expression of the training gene expression feature; Use the optimizer to perform backpropagation, calculate the gradient of the loss function on the parameters of the deep neural network that combines the diffusion model and the autoencoder, minimize the reconstruction error, and update the parameters of the encoder and the decoder; The deep neural network combining the diffusion model and the autoencoder is trained according to the number of cycles, and each batch performs the diffusion step, forward propagation, loss calculation, backpropagation and optimization.

5. The method for cancer prognosis assessment and target identification based on deep neural network according to claim 4, characterized in that: The optimizer is the Adam optimizer.

6. The method for cancer prognosis assessment and target identification based on deep neural network according to claim 4, characterized in that: The step of acquiring experimental data and preprocessing the experimental data to obtain training data includes: Gene expression data and survival data of cancer patients were obtained from TCGA and GEO public data sets, and the missing values ​​in the obtained data were filled with the average to obtain training data.

7. The method for cancer prognosis assessment and target identification based on deep neural network according to claim 4, characterized in that: The training process of the XGBoost survival analysis model includes the following steps: The low-dimensional expression of the training gene expression feature is used as input to construct a data format suitable for survival analysis; Initialize an XGBoost survival analysis model, and use the low-dimensional expression and survival time of the training gene expression features as labels to train the XGBoost survival analysis model.

8. A cancer prognosis assessment and target identification system based on deep neural network, characterized in that: The system comprises: A data acquisition module, the data acquisition module is used to acquire gene expression data and pre-process the gene expression data to obtain initial data; A feature acquisition module, wherein the feature acquisition module is used to perform feature extraction processing on the initial data using a deep neural network combining a diffusion model and an autoencoder to obtain a low-dimensional expression of gene expression features; An evaluation result acquisition module, wherein the evaluation result acquisition module is used to input the low-dimensional expression of the gene expression feature into the XGBoost survival analysis model to obtain a risk value, and obtain a cancer prognosis evaluation result according to the risk value; A target acquisition module is used to perform differential expression analysis based on the risk value and the patient's gene expression to obtain cancer prognosis targets.

9. A cancer prognosis assessment and target identification device based on deep neural network, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the cancer prognosis assessment and target identification method based on deep neural network as described in any one of claims 1-7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, which, when called, are used to execute the cancer prognosis assessment and target identification method based on a deep neural network as described in any one of claims 1 to 7.

Images