A method and application for predicting the therapeutic sensitivity of BCL2 inhibitor venetoclax in acute myeloid leukemia

Abstract

The present invention discloses a method for predicting the therapeutic sensitivity of BCL2 inhibitor venetoclax in acute myeloid leukemia, and belongs to the field of biomedical technology. The method disclosed in the present invention establishes a model through sample amino acid metabolism, fatty acid metabolism and glycolysis levels, divides acute myeloid leukemia patients into four categories, and through drug sensitivity database analysis, determines that three categories of medium metabolism are sensitive to venetoclax targeted therapy, and four categories of high metabolism are resistant to venetoclax targeted therapy. The method can efficiently evaluate and identify patients who are sensitive or resistant to venetoclax, and according to the metabolic characteristics of the four categories of resistant patients, provide them with a medication regimen of venetoclax combined with etomoxer to improve the treatment effect, and provide a tool for analyzing the therapeutic sensitivity of acute myeloid leukemia patients to venetoclax and providing individualized treatment.

Description

Technical Field

[0001] The present invention belongs to the field of biomedicine technology and specifically relates to a method for predicting the therapeutic sensitivity of BCL2 inhibitor venetoclax in acute myeloid leukemia. Background Art

[0002] Acute myeloid leukemia (AML) is a hematopoietic tumor characterized by clonal proliferation, differentiation blockade, and inhibition of apoptosis of malignant stem or progenitor cells. AML has a poor clinical prognosis and a high mortality rate, with an overall 5-year survival rate of less than 50%. Standard chemotherapy is the main treatment, but it sometimes fails to achieve complete remission (CR), and relapse may occur after CR. Currently, a new targeted drug treatment regimen, the BCL2 inhibitor venetoclax, has been proposed, which has achieved good results, but drug resistance still exists. Therefore, predicting the sensitivity of acute myeloid leukemia patients to venetoclax by certain means is of great significance for the individualized and efficient treatment of patients.

[0003] In recent years, metabolic reprogramming has been recognized as a basic feature of tumors. The Warburg effect proposed earlier showed that cancer cells mainly use aerobic glycolysis to provide energy for rapid proliferation; recent studies have found that mitochondrial metabolism also plays a vital role in tumor growth, and venetoclax plays a tumor-killing role by targeting amino acid metabolism and then affecting mitochondrial metabolism. Fatty acid metabolism and glycolysis, as the raw material sources of mitochondrial metabolism, may also affect the sensitivity of venetoclax. Summary of the invention

[0004] In view of the above problems, the present invention provides a method for predicting the sensitivity of BCL2 inhibitor venetoclax in acute myeloid leukemia. At present, there is a certain drug resistance in patients with acute myeloid leukemia treated with venetoclax, and this drug resistance is largely related to the metabolic state in AML. The prediction method in this patent can distinguish sensitive and resistant patients by amino acid metabolism, fatty acid metabolism and glycolysis levels, and solve the problem of drug resistance based on the metabolic characteristics of resistant patients.

[0005] In order to achieve the above object, the present invention provides the following technical solutions.

[0006] The present invention provides a method for predicting the therapeutic sensitivity of BCL2 inhibitor venetoclax in acute myeloid leukemia, characterized in that the method comprises the following steps:

[0007] Step 1: Obtain a training sample data set from the VIZOME and AML databases;

[0008] Step 2: Download the gene sets related to fatty acid metabolism, amino acid metabolism and glycolysis from the GSEA official website, extract the gene expression data of the three gene sets of VIZOME and AML data from the gene expression RPKM data in step 1, and calculate the ssGSEA score of each gene set;

[0009] Step 3: Classify the VIZOME and AML data according to the high and low score characteristics of the three gene sets. Extract the scores of the three gene sets in step 2 and use K-means clustering to classify the training sample data set.

[0010] Step 4: Determine the level of oxidative phosphorylation metabolism in four categories. Use the ssGSEA algorithm to calculate the oxidative phosphorylation metabolism level score of each sample with the oxidative phosphorylation gene set obtained from the GSEA official website as the range. According to the relative strength of expression, the samples are defined as Category I - low metabolism, Category II - medium metabolism (mainly fatty acid metabolism and glycolysis), Category III - medium metabolism (mainly amino acid metabolism), and Category IV - high metabolism;

[0011] Step 5: Obtain venetoclax sensitivity data from VIZOME, the AML drug sensitivity database, compare the AUC values ​​of venetoclax in the four categories, and determine that category III-intermediate metabolism (mainly amino acid metabolism) is sensitive to venetoclax, and category I-low metabolism, category II-intermediate metabolism (mainly fatty acid metabolism and glycolysis), and category IV-high metabolism are resistant to venetoclax;

[0012] Step 6: Obtain the gene expression data of the sample to be tested, use the R language precompiled package sva package to de-batch the gene expression data of the sample to be tested and the gene expression data in the VIZOME and AML databases, calculate the fatty acid metabolism, amino acid metabolism and glycolysis scores of the sample to be tested, and then use the machine learning algorithm to classify the sample to be tested into the type corresponding to the VIZOME database according to the expression pattern of fatty acid metabolism, amino acid metabolism and glycolysis scores.

[0013] Furthermore, the training sample data set in step 1 includes gene expression RPKM data and clinical information data of each sample of an acute myeloid leukemia patient.

[0014] Furthermore, the parameters of K-means clustering in step 3 are set as follows: number of simulations = 100, distance = Euclidean distance, connection mode = average, and after clustering, the number of clusters is set to 4 according to the consistency data.

[0015] Furthermore, the machine learning algorithm used in step 6 is a support vector machine (SVM).

[0016] The present invention also provides a storage medium having a plurality of instructions stored thereon, characterized in that the instructions are suitable for being loaded and executed by a processor to implement the steps of the above-mentioned method for predicting the therapeutic sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia.

[0017] The present invention also provides a terminal, characterized in that it comprises: a processor and a storage medium communicatively connected to the processor, wherein the storage medium is suitable for storing multiple instructions; the processor is suitable for calling the instructions in the storage medium to execute the steps of implementing the above-mentioned method for predicting the therapeutic sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia.

[0018] The present invention also provides a prediction model for the treatment sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia, characterized in that it is constructed by the steps of the above-mentioned method for predicting the treatment sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia.

[0019] The present invention also provides a method and device for predicting the therapeutic sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia, characterized in that it comprises the above-mentioned prediction model for the therapeutic sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia.

[0020] Compared with the prior art, the present invention has the following beneficial effects.

[0021] 1. This invention is the first to use the patient's sequencing data to calculate the metabolic level, and cluster AML patients according to the three scores of fatty acid metabolism, amino acid metabolism and glycolysis metabolism, and divide the patients into 4 categories: Category 1 - low metabolism, Category 2 - medium metabolism (dominated by fatty acid metabolism and glycolysis), Category 3 - medium metabolism (dominated by amino acid metabolism), Category 4 - high metabolism.

[0022] 2. The present invention confirms for the first time that category III patients are relatively sensitive to venetoclax, while category I, II, and IV patients are relatively resistant to venetoclax.

[0023] 3. The present invention is the first to discover that the sensitivity of four types of patients to venetoclax can be improved by combining it with etomoxir, a fatty acid metabolism inhibitor.

[0024] 4. The method disclosed in the present invention can efficiently evaluate and identify patients who are sensitive or resistant to venetoclax, and provide them with a venetoclax combined with etomoxir medication regimen to improve the treatment effect based on the metabolic characteristics of four types of resistant patients, thereby providing a tool for analyzing the therapeutic sensitivity of acute myeloid leukemia patients to venetoclax and providing individualized treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1A flow chart of a method for predicting sensitivity to venetoclax in patients with acute myeloid leukemia.

[0026] Figure 2 It is a validation of predicted sensitivity and the effect of combined treatment in AML cell lines. Figure A shows the relative sensitivity of two cell lines to venetoclax, and Figure B shows that KASUMI-1 combined with etomoxir can improve the sensitivity of venetoclax.

[0027] Figure 3 This is a classification prediction chart of the sensitivity of acute myeloid leukemia patients to venetoclax in the TCGA database. DETAILED DESCRIPTION

[0028] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will contribute to the understanding of the present invention, but these embodiments are only for the purpose of illustrating the present invention, and the present invention is not limited to these contents. The operating methods in the embodiments are all conventional operating methods in the art.

[0029] Example 1.

[0030] A method for predicting the therapeutic sensitivity of BCL2 inhibitor venetoclax in acute myeloid leukemia, the specific steps are as follows ( Figure 1 ).

[0031] Step 1: Obtain a training sample data set from the VIZOME, AML database. The training sample data set includes 211 data of acute myeloid leukemia patients. The 211 data include gene expression RPKM data and clinical information data of each sample.

[0032] Step 2: Download the gene sets related to fatty acid metabolism, amino acid metabolism, and glycolysis from the GSEA official website, extract the gene expression data of the three gene sets of VIZOME and AML data from the gene expression RPKM data in step 1, and calculate the ssGSEA score of each gene set.

[0033] Step 3: Classify the VIZOME and AML data according to the high and low score characteristics of the three gene sets. Extract the scores of the three gene sets in step 2 and use K-means clustering to classify the training sample data set.

[0034] Step 4: Determine the levels of the four types of oxidative phosphorylation metabolism. Use the ssGSEA algorithm to calculate the oxidative phosphorylation metabolism level score of each sample with the oxidative phosphorylation gene set obtained from the GSEA official website as the range. According to the relative strength of expression, the samples are defined as Class I - low metabolism, Class II - medium metabolism (mainly fatty acid metabolism and glycolysis), Class III - medium metabolism (mainly amino acid metabolism), and Class IV - high metabolism.

[0035] Step 5: For venetoclax sensitivity analysis, venetoclax sensitivity data were obtained from VIZOME, AML drug sensitivity database, and the AUC values ​​of venetoclax were compared in the four categories to determine that category III-intermediate metabolism (mainly amino acid metabolism) was sensitive to venetoclax, while category I-low metabolism, category II-intermediate metabolism (mainly fatty acid metabolism and glycolysis), and category IV-high metabolism were resistant to venetoclax.

[0036] Step 6: Venetoclax sensitivity analysis of the samples to be tested, obtain the gene expression data of the samples to be tested, use the R language precompiled package sva package to de-batch the gene expression data of the samples to be tested and the gene expression data in the VIZOME and AML databases, calculate the fatty acid metabolism, amino acid metabolism and glycolysis scores of the samples to be tested, and then use the machine learning algorithm support vector machine (SVM) to classify the samples to be tested into the types corresponding to the VIZOME database according to the expression patterns of fatty acid metabolism, amino acid metabolism and glycolysis scores.

[0037] Example 2.

[0038] The gene expression data of GDSC database and VIZOME and AML databases were de-batched using the R language precompiled package sva, and the fatty acid metabolism, amino acid metabolism and glycolysis scores in the GDSC database were calculated. Then, the machine learning algorithm support vector machine (SVM) was used to divide the AML cell lines in the GDSC database into the same four categories as the VIZOME database according to the expression patterns of fatty acid metabolism, amino acid metabolism and glycolysis scores. By calculating the AML cell line samples in the GDSC database, it was found that the MOLM13 cell line belonged to category III and the KASUMI-1 cell line belonged to category IV; the IC50 of the two cell lines was detected in vitro, which verified that MOLM13-category III was relatively sensitive to venetoclax, while KASUMI-1-category IV was relatively resistant to venetoclax; and it was verified that the combined use of the fatty acid metabolism inhibitor etomoxir with KASUMI-1 could improve the sensitivity of venetoclax.

[0039] Figure 2 It is a validation of predicted sensitivity and the effectiveness of combined therapy in AML cell lines.

[0040] Example 3.

[0041] Predicting the sensitivity of acute myeloid leukemia patients to venetoclax in the TCGA database: Obtain the gene expression data of TCGA and AML, use the R language precompiled package sva package to de-batch the gene expression data of the test samples and the gene expression data in the VIZOME and AML databases, calculate the fatty acid metabolism, amino acid metabolism and glycolysis scores of the test samples, and then use the machine learning algorithm support vector machine (SVM) to classify the test samples into the types corresponding to the VIZOME database according to the expression patterns of fatty acid metabolism, amino acid metabolism and glycolysis scores.

[0042] Figure 3 This is a classification prediction chart of acute myeloid leukemia patients' sensitivity to venetoclax in the TCGA database. Among them, Class III-Medium Metabolism (mainly amino acid metabolism) is sensitive to venetoclax, Class I-Low Metabolism, Class II-Medium Metabolism (mainly fatty acid metabolism and glycolysis) and Class IV-High Metabolism are resistant to venetoclax.

[0043] The above is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any technician familiar with the present invention can make some changes or modify the technical contents suggested above into equivalent embodiments without departing from the technical scope of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the solution of the present invention.

Claims

1. A storage medium having a plurality of instructions stored thereon, characterized in that: The instructions are suitable for being loaded and executed by a processor to implement the steps of a method for predicting the therapeutic sensitivity of the BCL2 inhibitor venetoclax in acute myeloid leukemia: Step 1: Obtain a training sample data set from the VIZOME and AML databases; Step 2: Download the gene sets related to fatty acid metabolism, amino acid metabolism, and glycolysis from the GSEA official website, extract the gene expression data of the three gene sets of VIZOME and AML data from the gene expression RPKM data in step 1, and calculate the ssGSEA score of each gene set; Step 3: Classify the VIZOME and AML data according to the high and low score characteristics of the three gene sets, extract the scores of the three gene sets in step 2, and use K-means clustering to classify the training sample data set; Step 4: The levels of the four types of oxidative phosphorylation metabolism were determined. The oxidative phosphorylation metabolism level score of each sample was calculated using the ssGSEA algorithm with the oxidative phosphorylation gene set obtained from the GSEA official website as the range. According to the relative strength of expression, the samples were defined as Class I - low metabolism, Class II - medium metabolism for fatty acid metabolism and glycolysis, Class III - medium metabolism for amino acid metabolism, and Class IV - high metabolism; Step 5: Obtain venetoclax sensitivity data from VIZOME, the AML drug sensitivity database, compare the AUC values ​​of venetoclax in the four categories, and determine that the third category of amino acid metabolism-intermediate metabolism is sensitive to venetoclax, the first category-low metabolism, the second category of fatty acid metabolism and glycolysis-intermediate metabolism, and the fourth category-high metabolism are resistant to venetoclax; Step 6: Obtain the gene expression data of the sample to be tested, use the R language precompiled package sva package to de-batch the gene expression data of the sample to be tested and the gene expression data in the VIZOME and AML databases, calculate the fatty acid metabolism, amino acid metabolism and glycolysis scores of the sample to be tested, and then use the machine learning algorithm to classify the sample to be tested into the type corresponding to the VIZOME database according to the expression pattern of fatty acid metabolism, amino acid metabolism and glycolysis scores.

2. A storage medium according to claim 1, characterized in that: The training sample data set described in step 1 includes gene expression RPKM data and clinical information data of each sample of an acute myeloid leukemia patient.

3. A storage medium according to claim 1, characterized in that: The parameters of K-means clustering in step 3 are set as follows: number of simulations = 100, Euclidean distance, average connection mode, and after clustering, the number of clusters is set to 4 according to the consistency data.

4. A storage medium according to claim 1, characterized in that: The machine learning algorithm used in step 6 is a support vector machine (SVM).

5. A terminal, characterized in that: include: A processor, a storage medium communicatively connected to the processor, the storage medium being suitable for storing a plurality of instructions; The storage medium is the storage medium according to any one of claims 1 to 4; The processor is adapted to invoke instructions in the storage medium.

Images