Psoriasis biological targeting antibody treatment response prediction system and device based on machine learning

Abstract

The invention discloses a psoriasis biological targeting antibody treatment response prediction device based on machine learning, and the psoriasis treatment response prediction device takes the expression level and clinical data of the differential expression gene of a psoriasis patient as the input, and further realizes the accurate prediction of the drug treatment response result of the psoriasis patient. When the psoriasis treatment response prediction device is used for drug treatment response prediction, the average ROC-AUC value is larger than 0.9, and it is indicated that the psoriasis treatment response prediction device can achieve accurate drug treatment response.

Description

Technical Field

[0001] The present invention relates to the field of biomedicine technology, and in particular to a system and device for predicting psoriasis biological targeted antibody treatment response based on machine learning. Background Art

[0002] Psoriasis, also commonly known as "psoriasis," is a chronic, recurring immune-mediated inflammatory skin disease. Symptoms vary, including red patches of skin often covered with silvery-white scales, itching, burning, or pain, dryness, cracking, and even bleeding. These patches often manifest as localized or widespread scaly erythema or plaques. Some patients also experience joint pain, swelling, and nail abnormalities.

[0003] The pathological mechanism of psoriasis is complex, involving multiple aspects of genetics, environment and the immune system. The main purpose of treating psoriasis is to control the disease and alleviate the patient's symptoms. Current treatment strategies mainly include biological therapy, phototherapy and topical drug therapy, but due to significant differences in treatment responses among patients, it is difficult to customize the best treatment plan for each patient in clinical practice.

[0004] With the development of bioinformatics and genetics, researchers have attempted to predict psoriasis patients' response to specific treatments by analyzing genetic markers and expression patterns. However, traditional prediction methods are often limited to small-scale data sets and single algorithms, lacking accuracy and generalization capabilities. In addition, although many studies have been devoted to gene expression analysis in psoriasis, translating these research results into practical clinical tools remains a challenge.

[0005] Therefore, there is a need to develop a method that can integrate data to improve the prediction accuracy of psoriasis treatment response and provide guidance for clinical personalized treatment. Summary of the Invention

[0006] The purpose of the present invention is to overcome the above-mentioned deficiencies in the prior art and to provide a psoriasis treatment response prediction system and device based on machine learning.

[0007] The first object of the present invention is to provide a device for predicting the response to psoriasis biological targeted antibody treatment based on machine learning.

[0008] The second object of the present invention is to provide the application of the above-mentioned psoriasis biological targeting antibody treatment response prediction device in judging the drug treatment response of psoriasis patients for non-diagnostic purposes.

[0009] The third object of the present invention is to provide a psoriasis biological targeting antibody treatment response prediction system based on machine learning.

[0010] A fourth object of the present invention is to provide an information data processing terminal.

[0011] A fifth object of the present invention is to provide a computer-readable storage medium.

[0012] A sixth object of the present invention is to provide a computer program product.

[0013] In order to achieve the above object, the present invention is implemented through the following scheme:

[0014] A device for predicting the response to psoriasis biological targeted antibody therapy based on machine learning, comprising a data acquisition module, a prediction module, and a result output module;

[0015] The data acquisition module is used to obtain the expression levels and clinical data of differentially expressed genes of psoriasis patients; the clinical data include medical history and / or treatment response;

[0016] The prediction module uses the expression levels of differentially expressed genes and clinical data obtained by the data acquisition module as input to the trained machine learning model to obtain the drug treatment response results of psoriasis patients;

[0017] The trained machine learning model is a machine learning model trained using the expression levels and clinical data of differentially expressed genes in psoriasis patients with known drug treatment responses;

[0018] The result output module is used to output the drug treatment response results of the psoriasis patients obtained by the prediction module.

[0019] Preferably, the differentially expressed genes of psoriasis patients are obtained by the following method:

[0020] S11. Obtain genetic and clinical information of 116 psoriasis patients from three US multicenter phase III clinical trials, AMAGINE-1.2.3, in the transcriptome dataset GSE117468 from the NCBI database;

[0021] The genetic information is the gene sequence data of psoriasis patients. The RAW data of transcriptome of GSE117468 in GEO database was downloaded in tar format using the function getGEO in the R package GEOquery.

[0022] Clinical information included the patient ID number, medication use of the psoriasis patient (including treatment with 140 mg of brodalumab, treatment with 210 mg of brodalumab, treatment with placebo, or treatment with ustekinumab), skin lesion status (including lesional skin or non-lesional skin), sampling time (baseline, 4 weeks, and 12 weeks, baseline is the time without treatment, which is equivalent to the prediction time), Psoriasis Area Severity Index (PASI, including PASI at baseline, 4 weeks, and 12 weeks), and drug treatment response at 4 weeks and drug treatment response at 12 weeks;

[0023] Among the drug treatment responses, a PASI score improvement of 75% compared to the baseline PASI score was considered a positive response, otherwise it was considered a negative response. The clinical information of psoriasis patients was obtained from the corresponding author of GSE117468.

[0024] S12. Performing quality control on the genetic and clinical information of the psoriasis patients obtained in step S11, including: performing data cleaning using an R package, performing dimensionality reduction analysis using the prcomp function, removing psoriasis patients treated with ustekinumab, and psoriasis patients lost to follow-up at 4 or 12 weeks (lacking PASI at 4 and 12 weeks), and grouping the psoriasis patients according to the annotations into a brodalumab 140 mg treatment group, a brodalumab 210 mg treatment group, and a placebo treatment group, to obtain the genetic and clinical information of the psoriasis patients after quality control;

[0025] Standardizing the genetic information and clinical information of the psoriasis patients after quality control to obtain a standardized data set of psoriasis patients (including the standardized genetic information and clinical information of the psoriasis patients);

[0026] The genetic information is standardized as follows: the genetic information of psoriasis patients is directly aligned to ACTB (Gene ID: 60) and GAPDH (Gene ID: 2597) using a multiple sequence alignment algorithm to obtain the standardized genetic information of psoriasis patients;

[0027] Clinical information was standardized as follows: based on the Psoriasis Area Severity Index (PASI) of psoriasis patients, psoriasis patients whose PASI at 4 weeks or 12 weeks improved by 75% compared with the PASI at baseline were judged as having a positive treatment response at 4 weeks or a positive treatment response at 12 weeks, and positive responses were recorded as 1, and other responses were recorded as 0.

[0028] S13. Based on the standardized psoriasis patient dataset obtained in step S12, use the R package limma to perform gene differential expression analysis on the genetic information of psoriasis patients in the brodalumab-treated group and the placebo-treated group to obtain differentially expressed genes;

[0029] S14. Based on the differentially expressed genes obtained in step S13, the top 1500 up- and down-regulated genes were selected, and gene function annotation and enrichment analysis were performed using KEGG. The biological processes, pathways, and gene molecules related to the pathogenesis of psoriasis were identified (wherein the biological processes, pathways, and gene molecules are recorded in PMID: 26025581), and the differentially expressed genes in psoriasis patients were obtained.

[0030] Preferably, the medical history and / or treatment response includes the psoriasis patient's medication usage (including treatment with 140 mg of Brodalumab, treatment with 210 mg of Brodalumab, treatment with placebo, or treatment with other medications), skin lesion status (including lesional LS or non-lesional NL), sampling time, and Psoriasis Area Severity Index (PASI).

[0031] Preferably, the drug treatment response result of the psoriasis patient is the drug treatment response result of the psoriasis patient 4 weeks and / or 12 weeks after taking the drug.

[0032] More preferably, the medication is Brodalumab.

[0033] Preferably, the machine learning model in the prediction module is a machine learning model that integrates the Lasso regression method and the LightGBM algorithm.

[0034] The present invention also seeks to protect the use of any of the above-mentioned psoriasis biological targeting antibody treatment response prediction devices in determining the drug treatment response of psoriasis patients for non-diagnostic purposes.

[0035] Preferably, the psoriasis patient's drug treatment response is the psoriasis patient's drug treatment response 4 weeks and / or 12 weeks after medication.

[0036] More preferably, the psoriasis patient's drug treatment response is a response to Brodalumab.

[0037] The present invention also claims protection for a psoriasis biological targeted antibody treatment response prediction system based on machine learning, comprising an acquisition unit, a storage unit, and a processing unit, wherein the acquisition unit is used to obtain the expression levels and clinical data of differentially expressed genes of psoriasis patients; the clinical data includes medical history and / or treatment response;

[0038] The storage unit stores program instructions that can be executed by the processing unit;

[0039] The processing unit contains any of the above-mentioned psoriasis treatment response prediction devices;

[0040] When the program instructions are executed by the processing unit, the expression levels of the differentially expressed genes of the psoriasis patients and the clinical data obtained by the acquisition unit are input into the processing unit to obtain the drug treatment response results of the psoriasis patients.

[0041] The present invention also seeks to protect an information data processing terminal, comprising a memory and a processor, wherein the memory stores a computer program that can be executed on the processor;

[0042] When the computer program is executed by the processor, the processor executes any of the above-mentioned psoriasis biological targeting antibody treatment response prediction devices and / or the above-mentioned psoriasis biological targeting antibody treatment response prediction systems.

[0043] The present invention also seeks to protect a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned psoriasis biological targeting antibody treatment response prediction devices and / or the above-mentioned psoriasis biological targeting antibody treatment response prediction systems.

[0044] The present invention also seeks to protect a computer program product, which includes the above-mentioned computer-readable storage medium.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention provides a device for predicting psoriasis treatment response based on biologically targeted antibodies based on machine learning, comprising a data acquisition module, a prediction module, and a result output module. The data acquisition module is configured to obtain expression levels of differentially expressed genes and clinical data of psoriasis patients; the clinical data includes medical history and / or treatment response; the prediction module uses the expression levels of differentially expressed genes and clinical data obtained by the data acquisition module as input, inputting them into a trained machine learning model to obtain the drug treatment response results of the psoriasis patients; the trained machine learning model is a machine learning model trained using the expression levels of differentially expressed genes and clinical data of psoriasis patients with known drug treatment response results; the result output module is configured to output the drug treatment response results of the psoriasis patients obtained by the prediction module. The device for predicting psoriasis treatment response based on biologically targeted antibodies uses the expression levels of differentially expressed genes and clinical data of psoriasis patients as input, thereby accurately predicting the drug treatment response results of psoriasis patients; when the device for predicting drug treatment response is used to predict drug treatment response, the average ROC-AUC values ​​are all greater than 0.9, indicating that the device for predicting psoriasis treatment response can accurately predict drug treatment response. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of the process of the psoriasis biological targeted antibody treatment response prediction model based on machine learning in Example 1;

[0048] Figure 2 Graphs showing the predicted treatment response at 12 weeks for each psoriasis patient in the test set using the psoriasis biotargeted antibody treatment response prediction model in Example 2; A is a ROC-AUC curve diagram for the predicted treatment response of 140 mg of Brodalumab after 12 weeks of treatment; B is a confusion matrix diagram for the predicted treatment response of 140 mg of Brodalumab after 12 weeks of treatment; C is a ROC-AUC curve diagram for the predicted treatment response of 210 mg of Brodalumab after 12 weeks of treatment; and D is a confusion matrix diagram for the predicted treatment response of 210 mg of Brodalumab after 12 weeks of treatment.

[0049] Figure 3Graphs showing the predicted treatment response at 4 weeks for each psoriasis patient in the test set using the psoriasis biotargeted antibody treatment response prediction model in Example 2; A is a ROC-AUC curve diagram of the treatment response prediction results for 140 mg of Brodalumab treatment for 4 weeks; B is a confusion matrix diagram of the treatment response prediction results for 140 mg of Brodalumab treatment for 4 weeks; C is a ROC-AUC curve diagram of the treatment response prediction results for 210 mg of Brodalumab treatment for 4 weeks; and D is a confusion matrix diagram of the treatment response prediction results for 210 mg of Brodalumab treatment for 4 weeks.

[0050] Figure 4 A schematic diagram of a non-volatile computer-readable storage medium for predicting response to psoriasis biotargeted antibody therapy. DETAILED DESCRIPTION

[0051] The present invention is further described in detail below with reference to the accompanying drawings and specific examples. The examples are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. The experimental methods used in the following examples are conventional methods unless otherwise specified; the materials and reagents used are commercially available unless otherwise specified.

[0052] In the embodiments of the present invention, before using the data of psoriasis patients for model training and verification, the patients are informed and their authorization is obtained.

[0053] Example 1 A machine learning-based prediction model for psoriasis biotargeted antibody treatment response

[0054] The process diagram of the psoriasis biotargeted antibody treatment response prediction model based on machine learning is as follows Figure 1 As shown, the details are as follows:

[0055] A machine learning-based prediction model for psoriasis biotargeted antibody treatment response, comprising a data acquisition module, a prediction module, and a result output module;

[0056] The data acquisition module is used to obtain the expression levels of differentially expressed genes in skin tissue samples of psoriasis patients and clinical data of psoriasis patients;

[0057] Clinical data included the psoriasis patients' medication use (treated with 140 mg of brodalumab, 210 mg of brodalumab, or placebo), lesional status (lesional LS or non-lesional NL), and sampling time (baseline (i.e., prediction time), 4 weeks, or 12 weeks);

[0058] The prediction module uses the expression levels of differentially expressed genes and clinical data obtained by the data acquisition module as input to the trained machine learning model to obtain the treatment response results of psoriasis patients after 4 weeks of medication and 12 weeks of medication;

[0059] The method for constructing a trained machine learning model specifically includes the following steps:

[0060] S1. Data Acquisition and Processing

[0061] S11. Obtain genetic and clinical information of 116 psoriasis patients from three US multicenter phase III clinical trials, AMAGINE-1.2.3, in the transcriptome dataset GSE117468 from the NCBI database;

[0062] The genetic information is the gene sequence data of psoriasis patients. The RAW data of transcriptome of GSE117468 in GEO database was downloaded in tar format using the function getGEO in the R package GEOquery.

[0063] Clinical information included the psoriasis patient ID, medication use (including treatment with 140 mg of brodalumab, 210 mg of brodalumab, placebo, or ustekinumab), lesion status (including lesional LS or non-lesional NL), sampling time (baseline, 4 weeks, and 12 weeks), Psoriasis Area Severity Index (PASI, including PASI at baseline, 4 weeks, and 12 weeks), and drug treatment response at 4 weeks and 12 weeks;

[0064] Among the drug treatment responses, a PASI score improvement of 75% compared to the baseline PASI score was considered a positive response, otherwise it was considered a negative response. The clinical information of psoriasis patients was obtained from the corresponding author of GSE117468.

[0065] S12. Performing quality control on the genetic and clinical information of the psoriasis patients obtained in step S11, including: performing data cleaning using an R package, performing dimensionality reduction analysis using the prcomp function, removing psoriasis patients treated with ustekinumab, and psoriasis patients lost to follow-up at 4 or 12 weeks (lacking PASI at 4 and 12 weeks), and grouping the psoriasis patients according to the annotations into a brodalumab 140 mg treatment group, a brodalumab 210 mg treatment group, and a placebo treatment group, to obtain the genetic and clinical information of the psoriasis patients after quality control;

[0066] Standardizing the genetic information and clinical information of the psoriasis patients after quality control to obtain a standardized data set of psoriasis patients (including the standardized genetic information and clinical information of the psoriasis patients);

[0067] The genetic information is standardized as follows: the genetic information of psoriasis patients is directly aligned to ACTB (Gene ID: 60) and GAPDH (Gene ID: 2597) using a multiple sequence alignment algorithm to obtain the standardized genetic information of psoriasis patients;

[0068] Clinical information was standardized as follows: based on the Psoriasis Area Severity Index (PASI) of psoriasis patients, psoriasis patients whose PASI at 4 weeks or 12 weeks improved by 75% compared with the PASI at baseline were judged as having a positive treatment response at 4 weeks or a positive treatment response at 12 weeks, and positive responses were recorded as 1, and other responses were recorded as 0.

[0069] S2. Data feature extraction

[0070] S211. Based on the standardized psoriasis patient dataset obtained in step S12, use the R package limma to perform gene differential expression analysis on the genetic information of psoriasis patients in the brodalumab-treated group and the placebo-treated group to obtain differentially expressed genes;

[0071] S212. Based on the differentially expressed genes obtained in step S211, select the top 1500 up- and down-regulated genes, perform gene function annotation and enrichment analysis using KEGG, and identify biological processes, pathways, and gene molecules related to the pathogenesis of psoriasis (wherein the biological processes, pathways, and gene molecules are described in PMID: 26025581), thereby obtaining brodalumab-responsive genes.

[0072] S213. Based on the Brodalumab response-related genes obtained in step S212 and the standardized psoriasis patient dataset obtained in step S12, a dataset for machine learning is constructed; wherein the dataset for machine learning includes the Brodalumab response-related genes in the standardized psoriasis patients and the standardized clinical information of the corresponding psoriasis patients;

[0073] S22. Based on the dataset for machine learning obtained in step S213, perform feature selection on the clinical information and genetic information of the samples in the dataset using Lasso regression to obtain weights of the clinical information and genetic information; wherein the regularization parameter λ in the Lasso regression analysis is 0.0026;

[0074] S23. Feature fusion: The clinical information and genetic information of each psoriasis patient are aggregated to obtain the information of each psoriasis patient, which is the feature fused dataset.

[0075] S3. Machine Learning Training

[0076] S31. Using a stratified sampling method, the feature fusion dataset obtained in step S23 is divided into a training set and a test set in a ratio of 7:3, and normalized using the Z-score method to obtain a processed training set;

[0077] S32. Use the LightGBM algorithm as the core algorithm for the machine learning model, set the learning rate to 0.01, the number of trees to 2000, the tree depth to 8, the maximum leaf size to 24, and use early stopping during training.

[0078] The genetic information of the processed samples in the training set was used as input features, and the treatment response results at 4 weeks and the drug treatment response results at 12 weeks were used as prediction targets. The machine learning model was trained through a 5-fold cross-validation combined with the clinical information of the processed samples in the training set to obtain a trained machine learning model.

[0079] The result output module is used to output the treatment response prediction results of psoriasis patients after 4 weeks of medication (including the treatment response prediction results after 4 weeks of using 140 mg of Brodalumab and the treatment effect prediction results after 4 weeks of using 210 mg of Brodalumab) and the treatment response prediction results after 12 weeks of medication (including the treatment response prediction results after 12 weeks of using 140 mg of Brodalumab and the treatment effect prediction results after 12 weeks of using 210 mg of Brodalumab).

[0080] Example 2 Performance Evaluation of a Machine Learning-Based Psoriasis Biological Targeted Antibody Therapy Response Prediction Model

[0081] 1. Experimental Methods

[0082] Obtaining genetic information and clinical information of each psoriasis patient in the test set shown in Example 1;

[0083] The genetic information is the expression level of Brodalumab response-related genes in each psoriasis patient in the test set, and the acquisition of Brodalumab response-related genes is described in Example 1; the clinical information is the treatment response results at 4 weeks and 12 weeks of each psoriasis patient in the test set.

[0084] The expression levels of Brodalumab response-related genes of each psoriasis patient in the test set were used as input features and input into the psoriasis biological targeted antibody treatment response prediction model established in Example 1. The treatment response prediction results of each psoriasis patient in the test set at 4 weeks of medication and 12 weeks of medication were obtained, and the ROC-AUC curve and confusion matrix diagram were drawn based on the clinical information of each psoriasis patient in the test set.

[0085] 2. Experimental Results

[0086] When the psoriasis biological targeted antibody treatment response prediction model of Example 1 is used to predict the treatment response of each psoriasis patient in the test set, the treatment response prediction results at 12 weeks are shown in the figure below: Figure 2 As shown, Figure 2 A is the ROC-AUC curve of the treatment response prediction results of 140 mg Brodalumab treatment for 12 weeks. Figure 2 B is the confusion matrix diagram of the treatment response prediction results of 140 mg Brodalumab treatment for 12 weeks. Figure 2 C is the ROC-AUC curve of the treatment response prediction results of 210 mg Brodalumab treatment for 12 weeks, Figure 2 D is the confusion matrix diagram of the treatment response prediction results after 12 weeks of treatment with 210 mg of brodalumab.

[0087] The prediction results of treatment response at 4 weeks were as follows Figure 3 As shown, Figure 3 A is the ROC-AUC curve of the treatment response prediction results after 4 weeks of treatment with 140 mg of Brodalumab. Figure 3 B is the confusion matrix diagram of the treatment response prediction results after 4 weeks of treatment with 140 mg of Brodalumab. Figure 3 C is the ROC-AUC curve of the treatment response prediction results after 4 weeks of treatment with 210 mg of Brodalumab. Figure 3 D is the confusion matrix diagram of the treatment response prediction results after 4 weeks of treatment with 210 mg of Brodalumab.

[0088] The results showed that when the psoriasis biological targeted antibody treatment response prediction model constructed in Example 1 was used to predict the treatment response of the test set, the average ROC-AUC value for predicting the treatment response of 140 mg of Brodalumab after 12 weeks of treatment was 0.9514, the average ROC-AUC value for predicting the treatment response of 210 mg of Brodalumab after 12 weeks of treatment was 0.9281, the average ROC-AUC value for predicting the treatment response of 140 mg of Brodalumab after 4 weeks of treatment was 0.9870, and the average ROC-AUC value for predicting the treatment response of 210 mg of Brodalumab after 4 weeks of treatment was 0.9751. The average ROC-AUC values ​​were all greater than 0.9, indicating that the psoriasis biological targeted antibody treatment response prediction model constructed in Example 1 can accurately predict whether psoriasis patients respond to Brodalumab treatment, and the model has excellent generalization.

[0089] Example 3 A machine learning-based prediction system for psoriasis biotargeted antibody treatment response

[0090] A psoriasis treatment response prediction system based on machine learning, comprising an acquisition unit, a storage unit, and a processing unit;

[0091] The acquiring unit is used to acquire genetic information of psoriasis patients; the genetic information of psoriasis patients is the expression level of Brodalumab response-related genes in psoriasis patients; wherein the acquisition of Brodalumab response-related genes is as described in Example 1;

[0092] The storage unit stores program instructions that can be executed by the processing unit;

[0093] The processing unit contains the psoriasis biological targeting antibody treatment response prediction model shown in Example 1;

[0094] When the program instructions are executed by the processing unit, the genetic information of the psoriasis patient obtained by the acquisition unit is input into the processing unit to obtain the treatment response results of the psoriasis patient after 4 weeks of medication and after 12 weeks of medication.

[0095] Example 4 A non-volatile computer-readable storage medium for predicting responses to psoriasis biotargeted antibody therapy

[0096] A non-volatile computer-readable storage medium for predicting the response to psoriasis biological targeted antibody therapy, as shown in FIG. Figure 4 As shown, a computer program is stored thereon, and when the computer program is executed by a processor, the psoriasis biological targeting antibody treatment response prediction model shown in Example 1 is implemented.

[0097] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Those skilled in the art will readily appreciate that other variations or modifications may be made based on the above descriptions and concepts. It is not necessary and impossible to provide an exhaustive list of all possible implementations. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A device for predicting the response to psoriasis biological targeted antibody therapy based on machine learning, characterized in that: Including data acquisition module, prediction module and result output module; The data acquisition module is used to obtain the expression levels and clinical data of differentially expressed genes of psoriasis patients; the clinical data include medical history and / or treatment response; The prediction module uses the expression levels of differentially expressed genes and clinical data obtained by the data acquisition module as input to the trained machine learning model to obtain the drug treatment response results of psoriasis patients; The trained machine learning model is a machine learning model trained using the expression levels and clinical data of differentially expressed genes in psoriasis patients with known drug treatment responses; The result output module is used to output the drug treatment response results of the psoriasis patients obtained by the prediction module.

2. The device for predicting response to psoriasis biologically targeted antibody therapy according to claim 1, characterized in that: The drug treatment response results of the psoriasis patients are drug treatment response results of the psoriasis patients 4 weeks and / or 12 weeks after taking the medication.

3. The device for predicting the response to treatment of psoriasis with biologically targeted antibodies according to claim 1, wherein: The machine learning model in the prediction module is a machine learning model that integrates the Lasso regression method and the LightGBM algorithm.

4. Use of the psoriasis biological targeting antibody treatment response prediction device according to any one of claims 1 to 3 in determining the response of psoriasis patients to drug treatment for non-diagnostic purposes.

5. The use according to claim 4, characterized in that The psoriasis patient's drug treatment response is the psoriasis patient's drug treatment response 4 weeks and / or 12 weeks after medication.

6. The use according to claim 5, characterized in that The psoriasis patient's drug treatment response is a response to Brodalumab.

7. A psoriasis biological targeted antibody treatment response prediction system based on machine learning, comprising an acquisition unit, a storage unit and a processing unit, characterized in that: The acquisition unit is used to acquire the expression levels of differentially expressed genes and clinical data of psoriasis patients; the clinical data include medical history and / or treatment response; The storage unit stores program instructions that can be executed by the processing unit; The processing unit comprises the psoriasis biological targeting antibody treatment response prediction device according to any one of claims 1 to 3; When the program instructions are executed by the processing unit, the expression levels of the differentially expressed genes of the psoriasis patients and the clinical data obtained by the acquisition unit are input into the processing unit to obtain the drug treatment response results of the psoriasis patients.

8. An information data processing terminal, comprising a memory and a processor, characterized in that: The memory stores a computer program executable on the processor; When the computer program is executed by the processor, the processor executes the device for predicting the response to treatment of psoriasis with biologically targeted antibodies according to any one of claims 1 to 3 and / or the system for predicting the response to treatment of psoriasis with biologically targeted antibodies according to claim 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the device for predicting the response to treatment of psoriasis with a biologically targeted antibody according to any one of claims 1 to 3 and / or the system for predicting the response to treatment of psoriasis with a biologically targeted antibody according to claim 7 are implemented.

10. A computer program product, characterized in that The computer program product comprises the computer-readable storage medium of claim 9.

Images