Method for detecting drug resistance of acute myelogenous leukemia by WDR4 and intelligent detection kit

Abstract

The invention discloses a method for detecting drug resistance of acute myelogenous leukemia by WDR4 and an intelligent detection kit, relates to the technical field of acute myelogenous leukemia detection, and aims to solve the problem of poor drug resistance detection effect of acute myelogenous leukemia. Through linkage detection of WDR4 gene, protein expression and drug resistance phenotype, a gene-protein-function complete evidence chain is formed, limitation of a single index is avoided, drug resistance judgment reliability is improved, accurate targets are provided for analysis of acute myelogenous leukemia drug resistance mechanism, linkage detection of WDR4 gene, protein expression and drug resistance phenotype is realized, and drug resistance of acute myelogenous leukemia is improved. A'gene-protein-function 'complete evidence chain is formed, the limitation of a single index is avoided, the drug resistance judgment reliability is improved, and a precise target spot is provided for analyzing the drug resistance mechanism of acute myelogenous leukemia.

Description

Technical Field

[0001] This invention relates to the field of acute myeloid leukemia detection technology, specifically to a method and intelligent detection kit for detecting drug resistance in acute myeloid leukemia using WDR4. Background Technology

[0002] Current methods for detecting drug resistance in acute myeloid leukemia (AML) largely rely on single-phenotype analysis, lacking validation of gene-protein associations, resulting in insufficient accuracy and specificity. Furthermore, inconsistent sample processing procedures and loosely linked testing steps lead to poor reproducibility and significant errors. In addition, the lack of reliable molecular markers makes early prediction of drug resistance risk difficult, often delaying treatment adjustments and impacting efficacy and prognosis. Therefore, there is an urgent need for more systematic and precise detection methods and tools. Summary of the Invention

[0003] The purpose of this invention is to provide a method and intelligent detection kit for detecting drug resistance in acute myeloid leukemia (AML) using WDR4. By detecting WDR4 gene, protein expression, and drug resistance phenotype in a coordinated manner, a complete "gene-protein-function" evidence chain is formed, avoiding the limitations of single indicators, improving the reliability of drug resistance assessment, and providing a precise target for elucidating the drug resistance mechanism of AML. This invention also addresses the problems in existing technologies.

[0004] To achieve the above objectives, the present invention provides the following technical solution: Methods for detecting drug resistance in acute myeloid leukemia using WDR4 include: First, biological samples were collected from patients with acute myeloid leukemia (AML). These samples were then processed using a kit to obtain leukemia cell samples. The expression level of the WDR4 gene in the leukemia cell samples was detected. Next, the expression level of the WDR4 protein in the leukemia cell samples was detected. The drug resistance phenotype of the leukemia cell samples to commonly used chemotherapy drugs was detected. A drug resistance association analysis was performed to determine the correlation between WDR4 gene expression level, WDR4 protein expression level, and drug resistance phenotype. Based on the results of the drug resistance association analysis, the drug resistance status of AML patients was determined.

[0005] Preferably, biological samples are collected from patients with acute myeloid leukemia, and the collected samples are processed in a kit, including: First, bone marrow or peripheral blood samples are collected from patients with acute myeloid leukemia. After collection, the samples are placed in a kit for sample pretreatment. The sample preprocessing procedure is as follows: The collected samples were first centrifuged once, and the first centrifugation yielded the sample supernatant and cell precipitate. The first supernatant was filtered. The cell pellet was then subjected to an ice bath, followed by a second centrifugation to obtain a second supernatant. Finally, the secondary supernatant and the primary supernatant were combined; After merging, leukemia cell samples were obtained.

[0006] Preferably, the expression level of the WDR4 gene in leukemia cell samples is detected, including: Total RNA was extracted from the obtained leukemia cell samples; The concentration and purity of the extracted total RNA were determined. After the concentration and purity of total RNA were determined, reverse transcription synthesis was performed, and cDNA products were obtained after reverse transcription synthesis. Using the synthesized cDNA product as a template, WDR4 gene-specific primers and internal reference gene primers were used for amplification. During the amplification process, the kit monitored and recorded changes in fluorescence signals in real time and generated amplification curves. The Ct values ​​of the WDR4 gene and internal reference gene in each sample are automatically calculated based on the generated amplification curve. The Ct value is the number of cycles when the fluorescence signal reaches the set threshold. The Ct values ​​of the WDR4 gene and the internal reference gene were compared, and the relative expression level of the WDR4 gene was calculated and confirmed by combining the reference samples in the database.

[0007] Preferably, the expression level of WDR4 protein in leukemia cell samples is detected, including: Add pre-chilled cell lysis buffer to the obtained leukemia cell sample and lyse for 30 minutes. After lysis, place the sample in the kit and incubate on ice for 30 minutes. After incubation on ice, centrifuge for 20 minutes. After centrifugation, take out the supernatant and store it in the kit. After storage, the total protein solution is obtained. The concentration of total protein solution was determined using the BCA method. The procedure was as follows: standard protein solutions were obtained from a protein database and serially diluted to prepare standard solutions of different concentrations. Meanwhile, diluent was added to the total protein solution for dilution, resulting in a diluted protein solution. Add the standard solution and protein dilution solution to the well plate, respectively, and then add BCA working solution to the standard solution and protein dilution solution and mix well; Place the well plate after mixing into the kit, incubate at 37°C for 30 minutes, then remove and cool to room temperature to obtain the mixed protein sample; The absorbance value at a wavelength of 562 nm was measured using an ELISA reader. A standard curve was plotted based on the absorbance of the standard, and the protein concentration of the mixed protein sample was calculated. The mixed protein samples were subjected to electrophoresis. After electrophoresis, the membrane was transferred. After the transfer, the membrane was removed and rinsed 2-3 times with TBST buffer for 5 minutes each time. After rinsing, the membrane was blocked. After blocking, the blocking solution was discarded and the membrane was rinsed 3 times with TBST buffer for 10 minutes each time. After rinsing, remove the membrane and blot the surface buffer with filter paper. Spread ECL chemiluminescence reagent evenly on the membrane and react for 1-2 minutes. Then, place it in a chemiluminescence imager for exposure to obtain protein band images. Finally, image analysis tools were used to analyze the gray values ​​of the WDR4 protein band and the internal reference protein band, and the ratio of the gray value of the WDR4 protein band to that of the internal reference protein band was calculated. After the calculation, the relative expression level of WDR4 protein was obtained.

[0008] Preferably, the leukemia cell samples are tested for resistance phenotypes to commonly used chemotherapy drugs, including: Leukemia cell samples were resuspended in RPMI-1640 medium containing 10% fetal bovine serum and the cell concentration was adjusted to 1× 100 μL of the sample was inoculated into a well plate at 37°C and 5% concentration. The culture medium was pre-cultured in an incubator for 24 hours to obtain the culture medium; Commonly used chemotherapy drugs for acute myeloid leukemia are selected and diluted into different concentration gradients using culture medium. Culture medium was added sequentially to drugs of different concentration gradients, and blank control group and negative control group were set up at the same time. Each group was set up with 3 replicates and cultured for 48 hours. After the culture was completed, 10 μL CCK-8 reagent was added to each well, and the absorbance of each well at 450 nm was measured using a microplate reader. Finally, cell viability was calculated based on absorbance values. A dose-response curve was plotted with drug concentration on the x-axis and cell viability on the y-axis. The half-maximal inhibitory concentration of the drug on leukemia cells was determined from the curve. If the half-maximal inhibitory concentration (MCI) of a leukemia cell sample is higher than that of a sensitive cell line, it indicates that the leukemia cell sample is resistant to the chemotherapy drug.

[0009] Preferably, an association analysis of drug resistance is performed on the relationship between WDR4 gene expression level, WDR4 protein expression level, and drug resistance phenotype, including: The correlation data between WDR4 gene expression level, WDR4 protein expression level and drug resistance phenotype were collected and labeled with leukemia cell sample numbers and corresponding clinical information to obtain the analysis dataset. The datasets were subjected to correlation analysis, group comparison analysis, and multi-factor association analysis. Correlation analysis was conducted for each chemotherapy drug, analyzing the correlation between WDR4 gene expression level and half-maximal inhibitory concentration (IC50) and between WDR4 protein expression level and IC50, including scatter plots and statistical tests. The grouped comparative analysis involved grouping leukemia cell samples according to the expression levels of the WDR4 gene or WDR4 protein, comparing the differences in drug resistance phenotypes between the two groups, including calculating the mean half-maximal inhibitory concentration (MCIC) for each group and displaying the differences between the two groups using a bar chart; statistical methods were used to test whether the difference in the MIC between the two groups was statistically significant. If the MIC of the high-expression group was higher than that of the low-expression group, it indicated that high WDR4 expression was associated with enhanced drug resistance. Multivariate association analysis involves analyzing the synergistic relationships of data in the dataset, including regression analysis, with the half-maximal inhibitory concentration (IC50) as the dependent variable and WDR4 gene expression and WDR4 protein expression as independent variables, to analyze the combined effect of both on drug resistance. Finally, correlation analysis, group comparison analysis, and multivariate association analysis were cross-validated to confirm whether there was a significant association between WDR4 gene expression level, protein expression level and drug resistance phenotype in acute myeloid leukemia, as well as the direction and strength of the association.

[0010] Preferably, the drug resistance status of patients with acute myeloid leukemia is determined based on the results of drug resistance association analysis, including: Key information was extracted from the drug resistance association analysis results, including the direction and strength of the association between WDR4 gene expression level and WDR4 protein expression level and drug resistance phenotype. At the same time, the critical values ​​for dividing the high-expression group and the low-expression group in the grouping analysis were confirmed, as well as the half-maximal inhibitory concentration (ICP-C) values ​​of the sensitive cell lines. Next, confirm the patient's core test data, including the relative expression level of the WDR4 gene, the relative expression level of the WDR4 protein, and the half-maximal inhibitory concentration (IC50) values ​​for different chemotherapy drugs. The patient's core test data were matched with the critical values ​​and thresholds determined in the drug resistance association analysis results; After matching, preliminary drug resistance data for the patients were obtained, including whether the patients' WDR4 gene expression level was high or low; whether the patients' WDR4 protein expression level was high or low; and whether the patients' half-maximal inhibitory concentration (IC50) values ​​for different chemotherapeutic drugs were higher than those of sensitive cell lines. Preliminary drug resistance data is used to determine a patient's sensitivity and resistance to a certain drug. Specifically, if a patient has high expression levels of both the WDR4 gene and WDR4 protein, and the half-maximal inhibitory concentration (IC50) is higher than the sensitivity threshold, the patient is considered to be in the high resistance range. If a patient has high expression levels of either the WDR4 gene or the WDR4 protein, and the IC50 is higher than the sensitivity threshold, the patient is considered to be in the intermediate resistance range, based on the correlation strength. If a patient has low expression levels of both the WDR4 gene and the WDR4 protein, and the IC50 is lower than the sensitivity threshold, the patient is considered to be in the sensitive range. The results of the preliminary drug resistance assessment will be verified by combining them with the patient's clinical information, which includes previous treatment response, relapse status, and disease progression. If the preliminary drug resistance data indicates high or low resistance, and the patient has previously experienced poor efficacy or relapse when using the drug, then the drug resistance should be reassessed. If the preliminary drug resistance data indicates a sensitive range, and the patient has previously responded well to the drug, then the drug resistance result should be confirmed. If the preliminary drug resistance data contradicts the clinical situation, then the reliability of the test data and correlation analysis should be re-examined. Finally, based on the combined validation results, the patient's drug resistance status was confirmed for each commonly used chemotherapy drug.

[0011] Preferably, using the synthesized cDNA product as a template, amplification is performed using WDR4 gene-specific primers and internal reference gene primers. During the amplification process, the kit monitors and records changes in fluorescence signal in real time and generates an amplification curve. Based on the generated amplification curve, the Ct values ​​of the WDR4 gene and internal reference gene in each sample are automatically calculated. The Ct value is the cycle number at which the fluorescence signal reaches a set threshold, including: Using the synthesized cDNA product as a template, amplification was performed using WDR4 gene-specific primers and internal reference gene primers. During the amplification process, the acquisition window was dynamically adjusted based on the amplification stage to acquire fluorescence signals, and fluorescence acquisition signals at each amplification stage were obtained. Fluorescence acquisition signals at each amplification stage were analyzed using sliding window variance mutation detection to determine the fluctuation range of fluorescence signals. Based on the variance of the fluorescence signal fluctuation range, the Gaussian filter parameters are adaptively adjusted to reduce the noise of the fluorescence acquisition signals at each amplification stage, and the denoised fluorescence acquisition signals at each amplification stage are obtained. The fluorescence acquisition signals of each amplification stage after noise reduction are fitted to generate amplification curves; Based on the five-parameter dual asymptote model and the dynamic setting threshold of the amplification efficiency coefficient α, the basic Ct values ​​of the WDR4 gene and the internal reference gene in each sample were solved to obtain the basic Ct values ​​of the WDR4 gene and the internal reference gene in each sample. The first auxiliary Ct value and the second auxiliary Ct value of the WDR4 gene were obtained by extracting the local curvature maximum point of the ellipse fitting and the slope maximum point of the polynomial fitting from the WDR4 gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the WDR4 gene; The final Ct value of the WDR4 gene was obtained by fusion of the weights of the five-parameter double asymptote model error, the local error of ellipse fitting, and the polynomial fitting error. The first auxiliary Ct value and the second auxiliary Ct value of the internal reference gene are obtained by extracting the local curvature maximum point of ellipse fitting and the slope maximum point of polynomial fitting from the internal reference gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the internal reference gene; The final Ct value of the internal reference gene is obtained by fusion of the weights adjusted in reverse based on the five-parameter dual asymptote model error, the local error of ellipse fitting, and the polynomial fitting error.

[0012] Preferably, before analyzing the grayscale values ​​of the WDR4 protein band and the internal reference protein band using image analysis tools, the method further includes: screening the protein band images to obtain the target protein band image; The process of filtering protein band images to obtain target protein band images includes: The protein band image is denoised using a bilateral filtering algorithm to obtain a denoised protein band image; one of the denoised protein band images is randomly selected as the image to be processed. The image to be processed is divided into several independent swimlane images; Each individual swimlane image is converted to grayscale; The WDR4 region in each independent lane image is slid based on the first preset window, and peak detection is performed on each sliding window. When a peak appears in a first preset number of consecutive first windows, the first preset number of consecutive first windows are taken as the effective region of the WDR4 region. The size of the first preset window is larger than the size of the second preset window, and the first preset number is less than the second preset number. Based on the second preset window, the intrinsic parameter region in each independent lane image is slid, and peak detection is performed on each sliding window. When a second preset number of consecutive second windows all show peaks, the second preset number of consecutive second windows are taken as the effective region of the intrinsic parameter region. The distribution continuity of pixels in the effective regions of the WDR4 region and the internal reference region is determined respectively. If the number of effective regions in either region is greater than 1, the image quality of that lane is deemed unqualified and is not used as the target protein band image. For the determined swimlane image, the optical density corresponding to all pixels in the effective region of the WDR4 region and the effective region of the internal reference region are calculated respectively. Based on the optical density corresponding to all pixels in the effective region of the WDR4 region, the integrated optical density and signal-to-noise ratio of the effective region of the WDR4 region are determined. Based on the optical density corresponding to all pixels in the effective region of the internal reference region, the integrated optical density and signal-to-noise ratio of the effective region of the internal reference region are determined. The edge sharpness value of the effective region of the WDR4 region is calculated. When the signal-to-noise ratio of the WDR4 region of the image to be processed is greater than the first preset signal-to-noise ratio threshold, the signal-to-noise ratio of the internal reference region is greater than the second preset signal-to-noise ratio threshold, and the edge sharpness value of the WDR4 region is greater than the preset edge sharpness threshold, the image to be processed is taken as the target protein strip image.

[0013] The WDR4 intelligent detection kit for drug resistance in acute myeloid leukemia includes: The kit contains an inner cavity with a slot at one end. A centrifuge and a filter are installed within this cavity. The centrifuge automatically identifies the blood sample type and matches centrifugation parameters using a sample weight sensor and AI algorithm. It also monitors centrifugation speed and vibration frequency in real time, automatically stopping and triggering an alarm via the control panel if any abnormality occurs. The filter integrates a laser particle size sensor, analyzing the size of impurities in the filtrate in real time. If purity is insufficient, it automatically initiates secondary filtration or replaces the filter membrane. It also automatically adjusts the filtration pressure based on the supernatant viscosity and displays the filtration progress on the control panel. A cooling rod and a placement rod are installed in the slot within the inner cavity. The cooling rod uses semiconductor cooling technology, with a temperature sensor monitoring the temperature in real time and automatically adjusting the cooling power. The cooling rod is synchronized with the control panel, automatically timing according to the experimental steps. A control panel is located on the outer wall of the kit. The placement rod contains an optical sensor that monitors the incubation status of the samples in the well plate in real time. The control panel automatically collects operational data from each device and generates a raw data report using a built-in algorithm.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The method and intelligent detection kit for detecting drug resistance in acute myeloid leukemia (AML) provided by this invention can detect WDR4 gene, protein expression and drug resistance phenotype in a linked manner, forming a complete evidence chain of "gene-protein-function", avoiding the limitations of single indicators, improving the reliability of drug resistance judgment, and providing a precise target for elucidating the drug resistance mechanism of AML.

[0015] 2. The method and intelligent detection kit for detecting drug resistance in acute myeloid leukemia using WDR4 provided by this invention integrate centrifugation, filtration, temperature control and other components to achieve centralized operation of the entire sample processing process, reduce errors and contamination, ensure consistent detection conditions, and improve the consistency and repeatability of results.

[0016] 3. The WDR4 method and intelligent detection kit for detecting drug resistance in acute myeloid leukemia provided by this invention combine clinical information to verify drug resistance grading, accurately match medication through correlation analysis, realize individualized treatment, improve treatment response rate, and avoid the toxic side effects and waste of resources of ineffective treatment. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the steps of the WDR4 method for detecting drug resistance in acute myeloid leukemia according to the present invention; Figure 2 This is a schematic diagram of the method for detecting drug resistance in acute myeloid leukemia using WDR4 according to the present invention.

[0018] In the diagram: 11. Inner cavity of the box; 12. Inner cavity slot; 13. Centrifuge; 14. Filter; 15. Cooling rod; 16. Placement rod; 17. Control panel. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] To address the problems in existing technologies regarding acute myeloid leukemia drug resistance testing—namely, the fragmented testing process, lack of integrated tools, and unsystematic association analysis between WDR4 and drug resistance, leading to low testing efficiency, poor result consistency, and difficulty in accurately guiding treatment—please refer to [the relevant documentation / reference]. Figure 1 and Figure 2 This embodiment provides the following technical solution: Methods for detecting drug resistance in acute myeloid leukemia using WDR4 include: First, biological samples were collected from patients with acute myeloid leukemia (AML). These samples were then processed using a kit to obtain leukemia cell samples. The expression level of the WDR4 gene in the leukemia cell samples was detected. Next, the expression level of the WDR4 protein in the leukemia cell samples was detected. The drug resistance phenotype of the leukemia cell samples to commonly used chemotherapy drugs was detected. A drug resistance association analysis was performed to determine the correlation between WDR4 gene expression level, WDR4 protein expression level, and drug resistance phenotype. Based on the results of the drug resistance association analysis, the drug resistance status of AML patients was determined.

[0021] Specifically, this method simultaneously detects WDR4 gene expression, protein expression, and drug resistance phenotype, forming a complete "gene-protein-function" evidence chain. Gene expression reflects transcriptional regulation, protein expression is directly related to functional activity, and the drug resistance phenotype reflects the actual drug response. The linkage of these three factors avoids the limitations of single indicators, significantly improving the reliability of drug resistance assessment. By clarifying the association between these factors and the drug resistance phenotype, the drug resistance mechanism mediated by WDR4 can be specifically identified, reducing false positive / false negative results and providing precise targets for elucidating the molecular mechanisms of AML drug resistance. Pre-treatment WDR4 expression detection can predict the risk of drug resistance in patients before or early in treatment, providing a basis for timely adjustment of treatment plans and avoiding the toxic side effects and resource waste caused by ineffective treatment. After clarifying the correspondence between WDR4 expression levels and resistance to specific chemotherapy drugs through correlation analysis, sensitive drugs can be selected "tailor-made" according to the patient's WDR4 expression characteristics, improving the treatment response rate. Furthermore, the detection technology used in this method is mature and stable, easy to promote and apply in clinical laboratories, and possesses both research value and clinical practicality.

[0022] The WDR4 intelligent detection kit for drug resistance in acute myeloid leukemia includes: The reagent kit has an inner cavity 11, with an inner cavity slot 12 at one end. A centrifuge 13 and a filter 14 are installed within the inner cavity 11. The centrifuge 13 connects to the AI ​​via a sample weight sensor. The algorithm automatically identifies the blood sample type and matches the centrifugation parameters. Simultaneously, it monitors the centrifugation speed and vibration frequency in real time. If any abnormality occurs, it automatically stops the centrifuge and alarms via the control panel 17. The filter 14 integrates a laser particle size sensor, analyzing the size of impurity particles in the filtrate in real time after filtration. If the purity is insufficient, it automatically initiates secondary filtration or replaces the filter membrane. Simultaneously, it automatically adjusts the filtration pressure based on the supernatant viscosity and displays the filtration progress via the control panel 17. A cooling rod 15 and a placement rod 16 are installed in the inner cavity slot 12. The cooling rod 15 uses semiconductor refrigeration technology, monitoring the temperature in real time with a temperature sensor and automatically adjusting the refrigeration power. The cooling rod 15 is synchronized with the control panel 17, automatically timing according to the experimental steps. The control panel 17 is located on the outer wall of the reagent kit. The placement rod 16 has a built-in optical sensor that monitors the incubation status of the samples in the well plate in real time. The control panel 17 automatically collects the operating data of each device and generates a raw data report using the built-in algorithm.

[0023] Specifically, the reagent kit features a hinged lid for easy opening and closing. The internal cavity integrates core components such as a centrifuge 13 and a filter 14. A slot 12 within the cavity specifically houses a cooling rod 15 and a placement rod 16. This concentrates key steps such as sample processing (centrifugation, filtration) and temperature control (ice bath, isothermal) into a single device, avoiding the cumbersome handling of multiple devices and sample transfers required in traditional testing. This reduces errors and contamination risks caused by operational steps, significantly improving the continuity of testing. The centrifuge is directly used for primary / secondary centrifugation of samples, the filter is adapted for supernatant purification, the cooling rod provides an ice bath environment to protect cell viability and biomolecular stability, and the isothermal rod meets the temperature control requirements for sample preservation (e.g., cDNA -20℃ storage) and reaction incubation (e.g., BCA method incubation at 37℃). The functions of each component are consistent with the detection methods. The sample processing and gene / protein detection steps are precisely matched, achieving equipment compatibility across the entire "sampling-processing-detection" process and ensuring consistency of experimental conditions. The outer control panel 17 can integrate parameter adjustment functions (such as centrifugation speed, cooling / constant temperature, reaction time, etc.), and achieves automated operation through preset programs, reducing human operation differences (such as centrifugation time control, temperature fluctuations), ensuring uniform conditions for different operators and different batches of tests, and improving the reproducibility and comparability of results. The partitioned design of the inner cavity slot 12 allows the cooling rod and the constant temperature rod to be placed independently, avoiding mutual temperature interference. The compact layout of the core equipment in the inner cavity 11 of the box saves experimental table space, which is especially suitable for the limited environment of clinical laboratories or small testing institutions. At the same time, it facilitates the overall storage and movement of the reagent kit, and has a certain degree of portability.

[0024] To address the issues in existing technologies where improper processing of acute myeloid leukemia samples, lack of standardization in WDR4 gene and protein expression detection procedures, and poor workflow integration lead to low sample purity, unreliable results, and negatively impact subsequent analyses, please refer to [the relevant documentation / reference]. Figure 1 and Figure 2 This embodiment provides the following technical solution: Biological samples were collected from patients with acute myeloid leukemia, and the collected samples were processed in kit 1, including: First, bone marrow or peripheral blood samples are collected from patients with acute myeloid leukemia. After collection, the samples are placed in kit 1 for sample pretreatment. The sample preprocessing procedure is as follows: First, centrifuge the collected samples once to obtain the primary supernatant and cell pellet. (Place the collected samples in centrifuge tube 13 and centrifuge at 1500-2000 r / min for 5-10 minutes.) The primary supernatant is filtered; (the primary supernatant in the centrifuge tube is taken out and placed into filter 14, which is filtered through a 0.22μm filter membrane to remove macromolecular impurities and cell debris.) The cell pellet was then subjected to an ice bath, followed by a second centrifugation to obtain a secondary supernatant. (The cell pellet from the centrifuge tube was placed in placer 16, and lysis buffer containing Tris-HCl, EDTA, and NP-40 was added. The mixture was lysed for 30 minutes in an ice bath using cooling rod 15. The sample was then removed from placer 16 and placed in centrifuge 13, where it was centrifuged at 12,000 rpm for 20 minutes to obtain a secondary supernatant.) Finally, the secondary supernatant and the primary supernatant are combined; (the secondary supernatant and the primary supernatant are combined in placer 16) After merging, leukemia cell samples were obtained.

[0025] Specifically, bone marrow or peripheral blood samples are selected as the collection targets, both of which are typical sources of AML cell enrichment: bone marrow is the core site of leukemia cell origin and proliferation, while peripheral blood often shows leukemia cell infiltration due to disease progression. This dual selection covers the routine testing needs of newly diagnosed patients and is also applicable to late-stage patients who cannot obtain bone marrow samples, improving the clinical universality of the protocol. The kit (1) is used for pretreatment, and the differences in human operation are reduced by standardized reagents and operating procedures, ensuring that the processing results between different laboratories and different operators are comparable, providing a stable sample basis for subsequent WDR4 expression detection and drug resistance analysis. The first centrifugation initially separates the supernatant and cell pellet, which can remove free proteins, metabolites and other interfering components in the plasma; the supernatant filtration further removes small cell debris and impurities, avoiding their interference with subsequent cell separation; the cell pellet is enriched by a second centrifugation, which can minimize non-target cells. The presence of residual target cells (such as normal hematopoietic cells) significantly improves the purity of leukemia cells, laying the foundation for accurate detection. The cell pellet is treated with an ice bath, which inhibits intracellular protease activity through a low-temperature environment, reduces cell structure degradation and damage to biomolecules (such as RNA and proteins), effectively maintains the physiological state and molecular stability of leukemia cells, and ensures the authenticity of subsequent WDR4 gene and protein expression detection results. By combining the primary and secondary supernatants, leukemia cells that may remain during centrifugation can be fully recovered, maximizing the use of the original sample. This is especially suitable for pediatric patients with small sample volumes or patients with difficult punctures, improving the utilization rate of limited samples and enhancing the practicality of the protocol.

[0026] The expression level of the WDR4 gene in leukemia cell samples was detected, including: Total RNA was extracted from the obtained leukemia cell samples; (the total RNA extraction procedure was as follows: Trizol reagent was added to the leukemia cell samples, with each 1× One leukemia cell sample corresponds to 1 mL of Trizol. The samples are thoroughly mixed to lyse the leukemia cells. Following the standard procedure in the reagent instructions, chloroform extraction and centrifugation are performed sequentially, collecting the upper aqueous phase containing RNA. Isopropanol is then added to precipitate the RNA, which is obtained by centrifugation. The precipitate is then washed with 75% ethanol to remove impurities. Finally, the ethanol is discarded by centrifugation. After the precipitate dries, it is dissolved in RNase-free ultrapure water to obtain the total RNA solution. The total RNA extracted was subjected to concentration and purity testing. (The procedure for concentration and purity testing was as follows: the extracted total RNA was tested using a UV spectrophotometer. The RNA solution was added to a cuvette, and the absorbance values ​​at wavelengths of 260 nm and 280 nm were measured. The ratio of the absorbance values ​​at 260 nm to 280 nm was used to assess RNA purity. The ratio of the absorbance values ​​at 260 nm to 280 nm was between 1.8 and 2.0. At the same time, the RNA concentration was calculated based on the absorbance value at 260 nm.) After the concentration and purity of total RNA were determined, reverse transcription synthesis was performed. After reverse transcription synthesis, cDNA product was obtained. (The reverse transcription synthesis procedure is as follows: add total RNA template, reverse transcriptase, primers, dNTP mixture and buffer to the reaction tube in sequence, mix gently, place the reaction tube in placer 16, and react according to the reaction conditions recommended in kit 1 to finally obtain cDNA product. The cDNA product is stored at -20℃.) Using the synthesized cDNA product as a template, WDR4 gene-specific primers and internal reference gene primers were used for amplification. During the amplification process, the kit (1) monitored and recorded the changes in fluorescence signal in real time and generated an amplification curve. (The amplification conditions were: 95℃ pre-denaturation for 30 seconds, then 95℃ denaturation for 5 seconds, 60℃ annealing for 30 seconds, for a total of 40 cycles. The WDR4 gene-specific primers and internal reference gene primers were GAPDH and β-actin.) The Ct values ​​of the WDR4 gene and internal reference gene in each sample are automatically calculated based on the generated amplification curve. The Ct value is the number of cycles when the fluorescence signal reaches the set threshold. The Ct values ​​of the WDR4 gene and the internal reference gene were compared, and the relative expression level of the WDR4 gene was calculated and confirmed by combining the reference samples in the database.

[0027] Specifically, Trizol reagent was used and the 1× A 1 mL reagent ratio per cell ensures thorough cell lysis and RNA release. The purification process, including chloroform extraction to separate the aqueous phase, isopropanol precipitation of RNA, and washing with 75% ethanol, forms a complete purification chain, effectively removing impurities such as proteins and genomic DNA. The resulting total RNA is highly pure and intact, providing a reliable foundation for subsequent experiments. The absorbance ratio at 260 nm / 280 nm (1.8-2.0) is measured using a UV spectrophotometer to precisely screen qualified RNA samples and avoid interference from protein contamination. Furthermore, the concentration is calculated based on the 260 nm absorbance, allowing for precise control of the amount of template used in subsequent reverse transcription, preventing errors due to excessive or insufficient template. To minimize errors and ensure experimental reliability from the outset, the reagent addition order is standardized and reaction conditions are strictly followed according to the kit recommendations, guaranteeing uniform reverse transcription efficiency. Storing cDNA products at -20℃ effectively inhibits nuclease activity, reduces degradation, ensures the reproducibility of subsequent amplification experiments, and minimizes deviations during sample processing. Using WDR4 gene-specific primers, combined with GAPDH and β-actin dual internal references for correction, effectively eliminates interference from differences in RNA extraction efficiency and reverse transcription between samples. The 40-cycle amplification conditions (95℃ pre-denaturation, denaturation, and 60℃ annealing) have been optimized to ensure amplification efficiency while reducing non-specific amplification. Real-time fluorescence monitoring generates amplification curves that visually reflect the amplification process, more accurately capturing gene expression differences compared to endpoint PCR, thus improving detection sensitivity. Ct value difference analysis combined with database reference samples enables standardized calculation of the relative expression level of the WDR4 gene, making the detection results of different samples and batches comparable and providing accurate quantitative data for subsequent drug resistance association analysis. This approach combines scientific rigor with practicality. The WDR4 gene expression level detection flowchart is shown below:

[0028] The expression level of WDR4 protein in leukemia cell samples was detected, including: Add pre-cooled cell lysis buffer to the obtained leukemia cell sample and lyse for 30 minutes. After lysis, place the sample on the cooling rod 15 of kit 1 and incubate on ice for 30 minutes. After incubation on ice, centrifuge for 20 minutes using centrifuge 13. After centrifugation, take out the supernatant and place it in the storage rod 16 for storage. After storage, the total protein solution is obtained. The concentration of total protein solution was determined using the BCA method. The procedure was as follows: standard protein solutions were obtained from a protein database and serially diluted to prepare standard solutions of different concentrations. Meanwhile, diluent was added to the total protein solution for dilution, resulting in a diluted protein solution. Add the standard solution and protein dilution solution to the well plate, respectively, and then add BCA working solution to the standard solution and protein dilution solution and mix well; Place the well plate after mixing into the placement rod 16, incubate at 37°C for 30 minutes, then remove and cool to room temperature to obtain the mixed protein sample. The absorbance value at a wavelength of 562 nm was measured using an ELISA reader. A standard curve was plotted based on the absorbance of the standard, and the protein concentration of the mixed protein sample was calculated. The mixed protein samples were subjected to electrophoresis. After electrophoresis, the membrane was transferred to a new membrane. After transfer, the membrane was removed and rinsed 2-3 times with TBST buffer for 5 minutes each time. After rinsing, the membrane was blocked. After blocking, the blocking solution was discarded, and the membrane was rinsed 3 times with TBST buffer for 10 minutes each time. (The electrophoresis process is as follows: Prepare a separating gel and a stacking gel of appropriate concentration according to the molecular weight of the target protein WDR4. Add an appropriate amount of loading buffer to the total protein solution, mix thoroughly, and heat in a boiling water bath for 5-10 minutes to denature the protein. After the gel solidifies, install it in the electrophoresis tank, add electrophoresis buffer, and add the denatured protein sample to the sample wells of the gel according to the calculated loading amount (usually 20-50 μg of protein per well). At the same time, add the protein molecular weight standard. Turn on the power and perform stacking gel electrophoresis at 80V. After the protein sample enters the separating gel, adjust the voltage to 120V and continue electrophoresis until the bromophenol blue indicator reaches the bottom of the gel. Then turn off the power.) The transfer process involves removing the gel after electrophoresis and preparing a nitrocellulose membrane (NC membrane) or polyvinylidene fluoride membrane (PVDF membrane) of the same size as the gel, along with filter paper. Soak the gel, membrane, and filter paper in transfer buffer until fully wetted. Assemble the gel in the transfer holder in the order of "sponge-filter paper-gel-membrane-filter paper-sponge," placing the holder in the transfer tank, ensuring the membrane faces the positive electrode and the gel faces the negative electrode. Add transfer buffer, select appropriate transfer conditions based on the membrane type and protein molecular weight, and start the transfer apparatus to transfer the protein from the gel to the membrane. The blocking process is as follows: After transfer, remove the membrane and rinse it 2-3 times with TBST buffer, 5 minutes each time. Then, place the membrane in blocking buffer and incubate with gentle shaking on a shaker at room temperature for 1-2 hours, or overnight at 4°C, to block non-specific binding sites on the membrane and reduce background interference in subsequent experiments. After blocking, discard the blocking buffer and rinse the membrane 3 times with TBST buffer, 10 minutes each time. Following the instructions for the WDR4 protein-specific primary antibody, dilute the primary antibody to an appropriate concentration with blocking buffer. Place the membrane in the diluted primary antibody solution and incubate with gentle shaking on a shaker overnight at 4°C. After incubation, rinse the membrane 3 times with TBST buffer, 10 minutes each time, to remove unbound primary antibody. Select a horseradish peroxidase (HRP)-labeled secondary antibody corresponding to the source of the primary antibody, dilute it to an appropriate concentration with blocking buffer, and place the membrane in the diluted secondary antibody solution. Incubate with gentle shaking on a shaker at room temperature for 1-2 hours. After incubation, rinse the membrane 3 times with TBST buffer, 10 minutes each time, to completely remove unbound secondary antibody.

[0029] After rinsing, remove the membrane and blot the surface buffer with filter paper. Spread ECL chemiluminescence reagent evenly on the membrane and react for 1-2 minutes. Then, place it in a chemiluminescence imager for exposure to obtain protein band images. Finally, image analysis tools were used to analyze the gray values ​​of the WDR4 protein band and the internal reference protein band, and the ratio of the gray value of the WDR4 protein band to that of the internal reference protein band was calculated. After the calculation, the relative expression level of WDR4 protein was obtained.

[0030] Specifically, using pre-cooled lysis buffer combined with ice bath lysis inhibits protease activity and reduces protein degradation; centrifugation to separate the supernatant and constant temperature storage maximizes the preservation of the integrity and native conformation of total protein, providing a high-quality sample basis for subsequent detection; the BCA method uses serial dilution of standard proteins to plot a standard curve, accurately quantifying the total protein concentration and ensuring uniform electrophoresis loading (20-50 μg), avoiding band deviations caused by sample volume differences, thus controlling experimental errors from the source; the gel concentration is customized according to the WDR4 molecular weight; denaturation treatment depolymerizes the protein; and stepwise voltage adjustment (80V concentration, 120V separation) ensures precise protein separation according to molecular weight; during membrane transfer, the normal and normal properties of the membrane and gel are maintained. Optimized polarity assembly and conditions ensure efficient protein transfer to the membrane, minimizing loss. Multiple TBST washes effectively remove non-specific binders. A blocking step inhibits excess binding sites on the membrane. Primary antibody incubation at 4°C overnight enhances specific binding, while secondary antibody HRP labeling improves signal sensitivity, collectively reducing background noise and making the target band clearly discernible. ECL colorimetry combined with chemiluminescence imaging captures weak signals. Grayscale analysis, corrected with internal reference proteins, eliminates sample processing differences, making the calculation of WDR4 protein relative expression levels more comparable and reproducible. This provides accurate protein level data for subsequent drug resistance association analysis. The WDR4 protein expression detection flowchart is shown below:

[0031] To address the shortcomings of existing technologies in detecting drug resistance in acute myeloid leukemia (AML), such as the lack of reliable molecular markers, unsystematic testing procedures and correlation analyses, difficulty in accurately assessing patient drug resistance, and thus impacting the effective selection of treatment regimens, please refer to [link to relevant documentation]. Figure 1 and Figure 2 This embodiment provides the following technical solution: The study involved testing leukemia cell samples for resistance phenotypes to commonly used chemotherapy drugs, including: Leukemia cell samples were resuspended in RPMI-1640 medium containing 10% fetal bovine serum and the cell concentration was adjusted to 1× 100 μL of the sample was inoculated into a well plate at 37°C and 5% concentration. The culture medium was pre-cultured in an incubator for 24 hours to obtain the culture medium; Commonly used chemotherapy drugs for acute myeloid leukemia are selected and diluted into different concentration gradients using culture medium. Culture medium was added sequentially to drugs of different concentration gradients, and blank control group and negative control group were set up at the same time. Each group was set up with 3 replicates and cultured for 48 hours. After the culture was completed, 10 μL CCK-8 reagent was added to each well, and the absorbance of each well at 450 nm was measured using a microplate reader. Finally, cell viability was calculated based on absorbance values. A dose-response curve was plotted with drug concentration on the x-axis and cell viability on the y-axis. The half-maximal inhibitory concentration of the drug on leukemia cells was determined from the curve. If the half-maximal inhibitory concentration (MCI) of a leukemia cell sample is higher than that of a sensitive cell line, it indicates that the leukemia cell sample is resistant to the chemotherapy drug.

[0032] Specifically, cells were resuspended in RPMI-1640 medium containing 10% fetal bovine serum, and the concentration was strictly adjusted to 1× Cells were pre-cultured at a density of [number] cells / mL for 24 hours to ensure uniform logarithmic growth, reducing interference from initial state differences on drug response and providing a stable cell model for subsequent experiments. Setting up drug concentration gradients accurately reflects the dose-response relationship. The inclusion of blank and negative control groups eliminates interference from culture medium, serum, and other factors, clarifying the drug's specific effects. The design of three replicates per group reduces random error and improves data reliability through parallel experiments. The CCK-8 reagent directly reflects cell viability through metabolic activity, and the absorbance value at 450nm objectively quantifies cell viability, avoiding subjective judgment bias. The dose-response curve plotted based on this provides a clear view of drug action. The calculation of the half-maximal inhibitory concentration (IC50) provides a quantitative standard for drug resistance assessment. Using sensitive cell lines as a reference, drug resistance is directly determined by the relative levels of samples, making the standard clear and operable. This result can accurately correlate with WDR4 gene and protein expression levels, providing functional evidence for elucidating drug resistance mechanisms and providing direct evidence for clinical selection of sensitive chemotherapeutic drugs, thus aiding in personalized treatment decisions.

[0033] The association between WDR4 gene expression level, WDR4 protein expression level, and drug resistance phenotype was analyzed using drug resistance association analysis, including: The correlation data between WDR4 gene expression level, WDR4 protein expression level and drug resistance phenotype were collected and labeled with leukemia cell sample numbers and corresponding clinical information to obtain the analysis dataset. The datasets were subjected to correlation analysis, group comparison analysis, and multi-factor association analysis. Correlation analysis was conducted for each chemotherapy drug, analyzing the correlation between WDR4 gene expression level and half-maximal inhibitory concentration (IC50) and between WDR4 protein expression level and IC50, including scatter plots and statistical tests. The grouped comparative analysis involved grouping leukemia cell samples according to the expression levels of the WDR4 gene or WDR4 protein, comparing the differences in drug resistance phenotypes between the two groups, including calculating the mean half-maximal inhibitory concentration (MCIC) for each group and displaying the differences between the two groups using a bar chart; statistical methods were used to test whether the difference in the MIC between the two groups was statistically significant. If the MIC of the high-expression group was higher than that of the low-expression group, it indicated that high WDR4 expression was associated with enhanced drug resistance. Multivariate association analysis involves analyzing the synergistic relationships of data in the dataset, including regression analysis, with the half-maximal inhibitory concentration (IC50) as the dependent variable and WDR4 gene expression and WDR4 protein expression as independent variables, to analyze the combined effect of both on drug resistance. Finally, correlation analysis, group comparison analysis, and multivariate association analysis were cross-validated to confirm whether there was a significant association between WDR4 gene expression level, protein expression level and drug resistance phenotype in acute myeloid leukemia, as well as the direction and strength of the association.

[0034] Specifically, by labeling sample numbers and organizing clinical information, the dataset systematically integrates gene and protein expression levels with drug resistance phenotype data to ensure comprehensive analytical dimensions and avoid biased conclusions due to data fragmentation. This provides reliable data support for subsequent multi-level analyses. Correlation analysis, targeting single drugs, visually presents the association trend between WDR4 expression level and half-maximal inhibitory concentration (MCIC) through scatter plots and statistical tests. Group comparison analysis, grouping by expression level, uses bar charts and statistical tests to clarify the direct association between high expression and enhanced drug resistance. Multivariate regression analysis reveals the synergistic effect of gene and protein expression, breaking through the limitations of single-indicator analysis. Overcoming the limitations of single-method analysis, the three-layer analysis forms a chain of evidence from individual to synergistic levels. By cross-validating the results of the three analyses, potential biases of a single method can be eliminated, and the direction (such as positive correlation) and strength of the association can be accurately confirmed, avoiding false positive conclusions. This makes the conclusion that "high WDR4 expression is associated with enhanced drug resistance" more statistically convincing. It not only clarifies the association between WDR4 and drug resistance but also quantifies the strength of the association, providing a precise basis for subsequent research on drug resistance reversal targeting WDR4. At the same time, it provides a scientific reference for clinical practice to predict drug resistance risk and formulate individualized treatment plans by detecting WDR4 expression levels, thus possessing both scientific research and clinical translational value.

[0035] Based on the results of drug resistance association analysis, the drug resistance status of patients with acute myeloid leukemia was determined, including: Key information was extracted from the drug resistance association analysis results, including the direction and strength of the association between WDR4 gene expression level and WDR4 protein expression level and drug resistance phenotype. At the same time, the critical values ​​for dividing the high-expression group and the low-expression group in the grouping analysis were confirmed, as well as the half-maximal inhibitory concentration (ICP-C) values ​​of the sensitive cell lines. Next, confirm the patient's core test data, including the relative expression level of the WDR4 gene, the relative expression level of the WDR4 protein, and the half-maximal inhibitory concentration (IC50) values ​​for different chemotherapy drugs. The patient's core test data were matched with the critical values ​​and thresholds determined in the drug resistance association analysis results; After matching, preliminary drug resistance data for the patients were obtained, including whether the patients' WDR4 gene expression level was high or low; whether the patients' WDR4 protein expression level was high or low; and whether the patients' half-maximal inhibitory concentration (IC50) values ​​for different chemotherapeutic drugs were higher than those of sensitive cell lines. Preliminary drug resistance data is used to determine a patient's sensitivity and resistance to a certain drug. Specifically, if a patient has high expression levels of both the WDR4 gene and WDR4 protein, and the half-maximal inhibitory concentration (IC50) is higher than the sensitivity threshold, the patient is considered to be in the high resistance range. If a patient has high expression levels of either the WDR4 gene or the WDR4 protein, and the IC50 is higher than the sensitivity threshold, the patient is considered to be in the intermediate resistance range, based on the correlation strength. If a patient has low expression levels of both the WDR4 gene and the WDR4 protein, and the IC50 is lower than the sensitivity threshold, the patient is considered to be in the sensitive range. The results of the preliminary drug resistance assessment will be verified by combining them with the patient's clinical information, which includes previous treatment response, relapse status, and disease progression. If the preliminary drug resistance data indicates high or low resistance, and the patient has previously experienced poor efficacy or relapse when using the drug, then the drug resistance should be reassessed. If the preliminary drug resistance data indicates a sensitive range, and the patient has previously responded well to the drug, then the drug resistance result should be confirmed. If the preliminary drug resistance data contradicts the clinical situation, then the reliability of the test data and correlation analysis should be re-examined. Finally, based on the combined validation results, the patient's drug resistance status was confirmed for each commonly used chemotherapy drug.

[0036] Specifically, by extracting key information such as the direction and strength of the association, and combining critical values ​​with threshold matching, patient drug resistance is divided into three ranges: high, medium, and low, breaking through the simple binary judgment model of "drug resistance / sensitivity". In particular, by combining gene and protein expression levels with the half-maximal inhibitory concentration (ICP-C) value, quantitative grading of drug resistance is achieved, providing a more detailed basis for clinical decision-making. After matching the core detection data with the association analysis results, further verification is performed by combining clinical information such as the patient's previous treatment response and relapse status, forming a closed loop of "detection data - analysis model - clinical practice". This verification mechanism can effectively filter out detection errors or deviations caused by individual special circumstances. For contradictions between preliminary results and clinical conditions, a re-verification process is set up to systematically check from detection data to association analysis model to ensure the scientific nature of the final conclusion. Meanwhile, secondary resistance assessments are conducted for cases with high / low resistance ranges and inconsistent clinical efficacy, demonstrating the rigor and flexibility of the protocol, reducing the risk of misjudgment, and confirming the resistance status of each commonly used chemotherapy drug individually. This allows for the direct customization of medication plans for patients, timely replacement of highly resistant drugs, and priority selection of sensitive drugs, avoiding the toxic side effects and resource waste caused by ineffective treatment. It is especially valuable for adjusting treatment in relapsed patients, ultimately achieving the goal of improving treatment response rates and patient prognosis.

[0037] Using the synthesized cDNA product as a template, amplification was performed using WDR4 gene-specific primers and internal reference gene primers. During the amplification process, the kit monitored and recorded changes in fluorescence signal in real time and generated amplification curves. Based on the generated amplification curves, the Ct values ​​of the WDR4 gene and internal reference gene in each sample were automatically calculated. The Ct value is the cycle number at which the fluorescence signal reaches a set threshold, including: Using the synthesized cDNA product as a template, amplification was performed using WDR4 gene-specific primers and internal reference gene primers. During the amplification process, the acquisition window was dynamically adjusted based on the amplification stage to acquire fluorescence signals, and fluorescence acquisition signals at each amplification stage were obtained. Fluorescence acquisition signals at each amplification stage were analyzed using sliding window variance mutation detection to determine the fluctuation range of fluorescence signals. Based on the variance of the fluorescence signal fluctuation range, the Gaussian filter parameters are adaptively adjusted to reduce the noise of the fluorescence acquisition signals at each amplification stage, and the denoised fluorescence acquisition signals at each amplification stage are obtained. The fluorescence acquisition signals of each amplification stage after noise reduction are fitted to generate amplification curves; Based on the five-parameter dual asymptote model and the dynamic setting threshold of the amplification efficiency coefficient α, the basic Ct values ​​of the WDR4 gene and the internal reference gene in each sample were solved to obtain the basic Ct values ​​of the WDR4 gene and the internal reference gene in each sample. The first auxiliary Ct value and the second auxiliary Ct value of the WDR4 gene were obtained by extracting the local curvature maximum point of the ellipse fitting and the slope maximum point of the polynomial fitting from the WDR4 gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the WDR4 gene; The final Ct value of the WDR4 gene was obtained by fusion of the weights of the five-parameter double asymptote model error, the local error of ellipse fitting, and the polynomial fitting error. The first auxiliary Ct value and the second auxiliary Ct value of the internal reference gene are obtained by extracting the local curvature maximum point of ellipse fitting and the slope maximum point of polynomial fitting from the internal reference gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the internal reference gene; The final Ct value of the internal reference gene is obtained by fusion of the weights adjusted in reverse based on the five-parameter dual asymptote model error, the local error of ellipse fitting, and the polynomial fitting error.

[0038] In this embodiment, the amplification phase includes a baseline phase, an exponential phase, and a plateau phase. The acquisition window size is dynamically adjusted based on the different phases of the amplification phase. Taking the WDR4 gene as an example: Baseline phase (1-12 cycles): When WDR4 expression is low, the signal is close to the background. 10 sets of data (cycle number + fluorescence value + timestamp) are retained to cover the entire baseline. 10 sets of internal controls are collected simultaneously to ensure consistency. Exponential phase (13-28 cycles): WDR4 enters rapid amplification. The window shrinks to 8 sets to focus on key regions. Internal controls are adjusted synchronously. Plateau phase (29-40 cycles): WDR4 amplification slows down (the plateau is blurred when expression is low). The window is restored to 10 sets to capture fluorescence stability. Internal controls are based on synchronous adaptation. When new data enters, the earliest set is removed according to the timestamp to ensure that the data reflects the dynamics of the current phase.

[0039] In this embodiment, the fluorescence acquisition signals at each amplification stage are determined by detecting variance abrupt changes through a sliding window, which includes: constructing a 5-cycle sliding window and calculating the variance. WDR4 threshold: (Noise is more significant at low expression levels, resulting in a higher threshold), mark the "fluctuation range" (e.g., cycles 15-17); <0.25 marks the "fluctuation recovery point"; internal reference threshold: (Stable expression, lower threshold), <0.25 is marked as "fluctuation recovery point"; Fluctuation range (Small standard deviation, retain weak signals); Non-fluctuation range (Large standard deviation, noise suppression); Internal reference: fluctuation range Non-fluctuation range (Adapt to stable amplification); If the WDR4 fluctuation point is still >2 after filtering, repeat the filtering once (maximum 2 times) to avoid over-smoothing.

[0040] In this embodiment, the fluorescence acquisition signals of each amplification stage after noise reduction are fitted to generate amplification curves, including: staged adaptive fitting. Global segmented fitting of the WDR4 gene, baseline period (cycles 1-12): ( The baseline background fluorescence level of the WDR4 gene is represented by the mean of the first 5 cycles, adapting to baseline blurring. This indicates the baseline fluorescence decay rate of the WDR4 gene; , indicating the baseline decay rate of the WDR4 gene; low purity indicates . Small); Exponential period (13~28 cycles): ( This represents the initial fluorescence value of the WDR4 gene during the exponential phase, and is taken as the smoothing signal of the 13th cycle. , represents the exponential amplification rate of the WDR4 gene; a smaller value is taken for low expression. The number of cycles indicating the start of the WDR4 gene exponential phase (taken as twice the mean fluorescence first exceeding baseline); plateau phase (cycles 29-40): ( This represents the maximum fluorescence value of the WDR4 gene during the plateau phase, with a value ranging from 35 to 40 cycles as the mean. This indicates the degree of fluorescence decrease during the plateau phase of the WDR4 gene; This indicates the plateau saturation rate of the WDR4 gene; This indicates the cycle number at which the WDR4 gene plateaus, and is taken as the first cycle when fluorescence growth rate is <0.01 / cycle.

[0041] Internal reference gene: baseline period (cycles 1-8): ( This represents the baseline background fluorescence level of the internal reference gene, and is taken as the average of the first 4 cycles. Indicates the baseline fluorescence decay rate of the internal reference gene; Fixed (indicating the baseline decay rate of the internal reference gene); exponential phase (cycles 9-22): ( Fixed, stable amplification, indicating the exponential amplification rate of the internal reference gene; This indicates the initial fluorescence value of the internal reference gene during the index period; The index phase (number of cycles at the start of the exponential phase) indicates the number of cycles for the internal reference gene; the plateau phase (cycles 23-40): ( This represents the maximum fluorescence value of the internal reference gene during the plateau phase, with a value ranging from the mean of 28 to 40 cycles. This indicates the degree of fluorescence decrease during the plateau phase of the internal reference gene; This indicates the plateau saturation rate of the internal reference gene; (Indicates the number of cycles at the plateau phase of the internal reference gene); Smoothing transition: Force fluorescence deviation at the two-stage transition point to be <3%, to avoid curve breakage (e.g., WDR4 exponential phase). (The transition during the plateau period).

[0042] Local dual-model comparison: Extracting local data from the exponential period 6 cycles (WDR4: cycles 16-21; internal parameters: cycles 12-17); Ellipse fitting: equation Calculate the local error of ellipse fitting. ( , This represents a local amplification inflection point. This indicates the length of the semi-major axis of the ellipse; (Represents the length of the minor semi-axis of the ellipse); Third-order polynomial fitting: , Represents a constant term; Denotes the coefficient of the linear term; Represents the coefficient of the quadratic term; Denotes the coefficients of the cubic term, used to calculate the polynomial fitting error. WDR4 gene: if (Low expression is closer to an ellipse), use the elliptic results to correct the global exponential period parameters (e.g.) , Internal reference gene: if (Stable amplification is closer to a polynomial), correct using polynomial results. Determine global fitting parameters (e.g., WDR4). , , , , , , , Fixed parameters of internal references , (etc.); local fitting model and error: ellipse , , , , Third-order polynomials , , , , .

[0043] In this embodiment, the baseline Ct values ​​of the WDR4 gene and the internal reference gene in each sample are solved based on a five-parameter dual asymptote model and a dynamically set threshold for the amplification efficiency coefficient α. This yields the baseline Ct values ​​of the WDR4 gene and the internal reference gene in each sample, including: a five-parameter dual asymptote model: Parameter assignment: (Lower asymptote): WDR4 gene values ​​were taken as the mean after smoothing during the baseline period (cycles 1-12); internal reference values ​​were taken as the mean during the baseline period (cycles 1-8). (Upper asymptote): WDR4 gene values ​​were taken as the mean after smoothing during the plateau phase (cycles 29-40); internal reference values ​​were taken as the mean during the plateau phase (cycles 23-40). (Amplification rate): WDR4 gene extraction (For low expression, use a small value, such as 0.15); internal references should be... =0.35 (fixed). (Plateau Inflection Point): WDR4 Gene Extraction Internal reference gene extraction ; Represents the baseline CT value; dynamic threshold WDR4 gene: threshold = ( Amplification efficiency coefficient: low expression , in expression High expression , compensation threshold sensitivity; (Indicates baseline fluorescence fluctuation of the WDR4 gene); Internal reference gene: threshold = ( Fixed, adapted for stable amplification; (Indicates the baseline fluorescence fluctuation of the internal reference gene); calculate the baseline ct values ​​of the WDR4 gene and the internal reference gene. Substitute the threshold value to calculate the number of cycles when fluorescence reaches the threshold (retain two decimal places).

[0044] In this embodiment, the maximum local curvature point of the ellipse fitting and the maximum slope point of the polynomial fitting are extracted from the WDR4 gene to obtain the first auxiliary Ct value and the second auxiliary Ct value of the WDR4 gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the WDR4 gene; The final Ct value of the WDR4 gene was obtained by fusion of the weights adjusted in reverse based on the five-parameter dual asymptote model error, the local error of ellipse fitting, and the polynomial fitting error. The points of maximum curvature of the ellipse fitting and the points of maximum slope of the polynomial fitting were extracted from the internal reference gene to obtain the first auxiliary Ct value and the second auxiliary Ct value of the internal reference gene, including: Auxiliary Ct value: The first auxiliary Ct value of the WDR4 gene, i.e., the ellipse fitting auxiliary Ct ( ): Number of iterations to obtain the endpoints of the minor axis of the local ellipse fitting ( The corresponding cycle number reflects the steepest stage of local amplification; the second auxiliary Ct value of the WDR4 gene, i.e., the polynomial fitting auxiliary Ct ( Differentiating a third-order polynomial ,untie and The number of cycles at the point with the maximum slope (corresponding to the stage with the fastest amplification rate) is obtained. The first auxiliary Ct value of the internal reference gene is the ellipse fitting auxiliary Ct ( ): Number of iterations to obtain the endpoints of the minor axis of the local ellipse fitting ( The corresponding cycle number reflects the steepest stage of local amplification; the second auxiliary Ct value of the internal reference gene, i.e., the polynomial fitting auxiliary Ct ( Differentiating a third-order polynomial ,untie and The number of cycles at the point with the maximum slope (corresponding to the stage with the fastest amplification rate) is obtained. Screening rules: WDR4 gene: if and Bias > 1.5 cycles (common in low expression), remove. Only retain , Internal reference gene: if and If the bias is >0.8 cycles (the internal reference gene is stable, and the bias should be even smaller), remove the affected gene. After screening, only a few may remain. , .

[0045] In this embodiment, the final Ct value of the WDR4 gene is obtained by fusion of the weights of the five-parameter double asymptote model error, the local error of ellipse fitting, and the polynomial fitting error. The final Ct value of the internal reference gene is obtained by fusion of the weights adjusted in reverse based on the five-parameter dual asymptote model error, the local error of ellipse fitting, and the polynomial fitting error. This includes: WDR4 gene: (Five-parameter model) (Ellipse, if retained) ), (Polynomial fitting); Internal reference genes: calculated similarly. 、 、 ; Determine the fusion weights based on the weighting formula. (The smaller the error, the greater the weight) The number of valid Ct values, such as WDR4 culling. back When determined After that, =1- When determined After that, =1- Assumption =0.8; =1.2, then but =0.4; Final Ct value = (The same applies to internal references); WDR4 : Recalculate baseline (cycles 1-15, expanding the range) and threshold ( =1.6 (to improve sensitivity) and avoid missed detections; internal reference The message indicates "sample degradation" (the internal control should be highly expressed with a low Ct value).

[0046] The working principle and beneficial effects of the above technical solution are as follows: By dynamically adjusting the acquisition window to adapt to different amplification stages, and combining the sliding window variance mutation detection to locate the signal fluctuation range, and then adaptively adjusting the Gaussian filter parameters based on the interval variance to reduce noise, interference is effectively reduced, ensuring the accuracy and stability of the fluorescence signal, and providing a high-quality data foundation for the generation of amplification curves; the basic Ct value is determined by a five-parameter dual asymptote model combined with the amplification efficiency coefficient α, and auxiliary Ct values ​​are obtained through ellipse fitting and polynomial fitting. Furthermore, the weights are adjusted inversely based on the errors of the three calculation methods to fuse the final Ct value, avoiding the limitations of a single method, significantly reducing the Ct value calculation error, and improving the reliability of the results; a unified signal processing and Ct value calculation logic is used for the WDR4 gene and the internal reference gene to ensure the comparability of the two data, providing accurate and consistent basic data for subsequent correction of WDR4 gene expression levels through the internal reference.

[0047] Before using image analysis tools to analyze the grayscale values ​​of the WDR4 protein band and the internal reference protein band, the following steps are also included: screening the protein band images to obtain the target protein band image; The process of filtering protein band images to obtain target protein band images includes: The protein band image is denoised using a bilateral filtering algorithm to obtain a denoised protein band image; one of the denoised protein band images is randomly selected as the image to be processed. The image to be processed is divided into several independent swimlane images; Each individual swimlane image is converted to grayscale; The WDR4 region in each independent lane image is slid based on the first preset window, and peak detection is performed on each sliding window. When a peak appears in a first preset number of consecutive first windows, the first preset number of consecutive first windows are taken as the effective region of the WDR4 region. The size of the first preset window is larger than the size of the second preset window, and the first preset number is less than the second preset number. Based on the second preset window, the intrinsic parameter region in each independent lane image is slid, and peak detection is performed on each sliding window. When a second preset number of consecutive second windows all show peaks, the second preset number of consecutive second windows are taken as the effective region of the intrinsic parameter region. The distribution continuity of pixels in the effective regions of the WDR4 region and the internal reference region is determined respectively. If the number of effective regions in either region is greater than 1, the image quality of that lane is deemed unqualified and is not used as the target protein band image. For the determined swimlane image, the optical density corresponding to all pixels in the effective region of the WDR4 region and the effective region of the internal reference region are calculated respectively. Based on the optical density corresponding to all pixels in the effective region of the WDR4 region, the integrated optical density and signal-to-noise ratio of the effective region of the WDR4 region are determined. Based on the optical density corresponding to all pixels in the effective region of the internal reference region, the integrated optical density and signal-to-noise ratio of the effective region of the internal reference region are determined. The edge sharpness value of the effective region of the WDR4 region is calculated. When the signal-to-noise ratio of the WDR4 region of the image to be processed is greater than the first preset signal-to-noise ratio threshold, the signal-to-noise ratio of the internal reference region is greater than the second preset signal-to-noise ratio threshold, and the edge sharpness value of the WDR4 region is greater than the preset edge sharpness threshold, the image to be processed is taken as the target protein strip image.

[0048] In this embodiment, the optical density corresponding to all pixels in the effective region of the WDR4 region and the effective region of the intrinsic parameter region is calculated, including: in, This represents the optical density value of the i-th pixel; This represents the grayscale value of the i-th pixel; This represents the average background gray value calculated from the signal-free region surrounding the protein band.

[0049] In this embodiment, the effective region signal-to-noise ratio of the WDR4 region is calculated as the ratio of the average optical density of the effective region of the WDR4 region to the standard deviation of the background region of the WDR4 region; the signal-to-noise ratio of the effective region of the intrinsic reference region is calculated in the same way.

[0050] In this embodiment, the edge sharpness value of the effective region of WDR4 is calculated, which is the ratio of the maximum gray-level gradient of the central region of WDR4 to the width of the protein band.

[0051] In this embodiment, the second preset signal-to-noise ratio threshold is greater than the first preset signal-to-noise ratio threshold.

[0052] The working principle and beneficial effects of the above technical solution are as follows: By using bilateral filtering for noise reduction, interference noise in protein band images can be effectively removed, laying a clear image foundation for subsequent region identification and analysis, and reducing the impact of noise on subsequent detection results; By using preset windows of different sizes (first and second preset windows) and different numbers of continuous peak judgment criteria for the WDR4 region and the internal reference region, the two types of target regions can be accurately distinguished and located, avoiding analysis bias caused by region confusion; By judging the number of effective regions (excluding unqualified lanes with more than 1), verifying the signal-to-noise ratio (both WDR4 and internal reference regions meet the standard) and edge sharpness (WDR4 region meets the standard), target protein band images with qualified image quality can be strictly screened, and images with poor quality and large interference can be removed, ensuring the reliability of samples for subsequent grayscale value analysis; Qualified target images provide a prerequisite for the accurate analysis of grayscale values ​​of WDR4 protein and internal reference protein, reducing grayscale value calculation errors caused by image quality problems, and improving the accuracy of the final protein expression level analysis results.

[0053] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0054] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for detecting drug resistance in acute myeloid leukemia using WDR4, characterized in that, include: First, biological samples are collected from patients with acute myeloid leukemia, and then the collected samples are placed in a kit for processing to obtain leukemia cell samples. The expression level of the WDR4 gene in leukemia cell samples was detected, and then the expression level of the WDR4 protein in leukemia cell samples was detected. The resistance phenotype of leukemia cell samples to commonly used chemotherapy drugs was detected. Drug resistance association analysis was performed to determine the correlation between WDR4 gene expression level, WDR4 protein expression level, and drug resistance phenotype. Based on the results of the drug resistance association analysis, the drug resistance status of patients with acute myeloid leukemia was determined.

2. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 1, characterized in that, Biological samples were collected from patients with acute myeloid leukemia, and the collected samples were then processed in a kit, including: First, bone marrow or peripheral blood samples are collected from patients with acute myeloid leukemia. After collection, the samples are placed in a kit for sample pretreatment. The sample preprocessing procedure is as follows: The collected samples were first centrifuged once, and the first centrifugation yielded the sample supernatant and cell precipitate. The first supernatant was filtered. The cell pellet was then subjected to an ice bath, followed by a second centrifugation to obtain a second supernatant. Finally, the secondary supernatant and the primary supernatant were combined; After merging, leukemia cell samples were obtained.

3. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 2, characterized in that, The expression level of the WDR4 gene in leukemia cell samples was detected, including: Total RNA was extracted from the obtained leukemia cell samples; The concentration and purity of the extracted total RNA were determined. After the concentration and purity of total RNA were determined, reverse transcription synthesis was performed, and cDNA products were obtained after reverse transcription synthesis. Using the synthesized cDNA product as a template, WDR4 gene-specific primers and internal reference gene primers were used for amplification. During the amplification process, the kit monitored and recorded changes in fluorescence signals in real time and generated amplification curves. The Ct values ​​of the WDR4 gene and internal reference gene in each sample are automatically calculated based on the generated amplification curve. The Ct value is the number of cycles when the fluorescence signal reaches the set threshold. The Ct values ​​of the WDR4 gene and the internal reference gene were compared, and the relative expression level of the WDR4 gene was calculated and confirmed by combining the reference samples in the database.

4. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 3, characterized in that, The expression level of WDR4 protein in leukemia cell samples was detected, including: Add pre-chilled cell lysis buffer to the obtained leukemia cell sample and lyse for 30 minutes. After lysis, place the sample in the kit and incubate on ice for 30 minutes. After incubation on ice, centrifuge for 20 minutes. After centrifugation, take out the supernatant and store it in the kit. After storage, the total protein solution is obtained. The concentration of total protein solution was determined using the BCA method. The procedure was as follows: standard protein solutions were obtained from a protein database and serially diluted to prepare standard solutions of different concentrations. Meanwhile, diluent was added to the total protein solution for dilution, resulting in a diluted protein solution. Add the standard solution and protein dilution solution to the well plate, respectively, and then add BCA working solution to the standard solution and protein dilution solution and mix well; Place the well plate after mixing into the kit, incubate at 37°C for 30 minutes, then remove and cool to room temperature to obtain the mixed protein sample; The absorbance value at a wavelength of 562 nm was measured using an ELISA reader. A standard curve was plotted based on the absorbance of the standard, and the protein concentration of the mixed protein sample was calculated. The mixed protein samples were subjected to electrophoresis. After electrophoresis, the membrane was transferred. After the transfer, the membrane was removed and rinsed 2-3 times with TBST buffer for 5 minutes each time. After rinsing, the membrane was blocked. After blocking, the blocking solution was discarded and the membrane was rinsed 3 times with TBST buffer for 10 minutes each time. After rinsing, remove the membrane and blot the surface buffer with filter paper. Spread ECL chemiluminescence reagent evenly on the membrane and react for 1-2 minutes. Then, place it in a chemiluminescence imager for exposure to obtain protein band images. Finally, image analysis tools were used to analyze the gray values ​​of the WDR4 protein band and the internal reference protein band, and the ratio of the gray value of the WDR4 protein band to that of the internal reference protein band was calculated. After the calculation, the relative expression level of WDR4 protein was obtained.

5. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 4, characterized in that, The study involved testing leukemia cell samples for resistance phenotypes to commonly used chemotherapy drugs, including: Leukemia cell samples were resuspended in RPMI-1640 medium containing 10% fetal bovine serum and the cell concentration was adjusted to 1× 100 μL of the sample was inoculated into a well plate at 37°C and 5% concentration. The culture medium was pre-cultured in an incubator for 24 hours to obtain the culture medium; Commonly used chemotherapy drugs for acute myeloid leukemia are selected and diluted into different concentration gradients using culture medium. Culture medium was added sequentially to drugs of different concentration gradients, and blank control group and negative control group were set up at the same time. Each group was set up with 3 replicates and cultured for 48 hours. After the culture was completed, 10 μL CCK-8 reagent was added to each well, and the absorbance of each well at 450 nm was measured using a microplate reader. Finally, cell viability was calculated based on absorbance values. A dose-response curve was plotted with drug concentration on the x-axis and cell viability on the y-axis. The half-maximal inhibitory concentration of the drug on leukemia cells was determined from the curve. If the half-maximal inhibitory concentration (MCI) of a leukemia cell sample is higher than that of a sensitive cell line, it indicates that the leukemia cell sample is resistant to the chemotherapy drug.

6. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 5, characterized in that, The association between WDR4 gene expression level, WDR4 protein expression level, and drug resistance phenotype was analyzed using drug resistance association analysis, including: The correlation data between WDR4 gene expression level, WDR4 protein expression level and drug resistance phenotype were collected and labeled with leukemia cell sample numbers and corresponding clinical information to obtain the analysis dataset. The datasets were subjected to correlation analysis, group comparison analysis, and multi-factor association analysis. Correlation analysis was conducted for each chemotherapy drug, analyzing the correlation between WDR4 gene expression level and half-maximal inhibitory concentration (IC50) and between WDR4 protein expression level and IC50, including scatter plots and statistical tests. The grouped comparative analysis involved grouping leukemia cell samples according to the expression levels of the WDR4 gene or WDR4 protein, comparing the differences in drug resistance phenotypes between the two groups, including calculating the mean half-maximal inhibitory concentration (MCIC) for each group and displaying the differences between the two groups using a bar chart; statistical methods were used to test whether the difference in the MIC between the two groups was statistically significant. If the MIC of the high-expression group was higher than that of the low-expression group, it indicated that high WDR4 expression was associated with enhanced drug resistance. Multivariate association analysis involves analyzing the synergistic relationships of data in the dataset, including regression analysis, with the half-maximal inhibitory concentration (IC50) as the dependent variable and WDR4 gene expression and WDR4 protein expression as independent variables, to analyze the combined effect of both on drug resistance. Finally, correlation analysis, group comparison analysis, and multivariate association analysis were cross-validated to confirm whether there was a significant association between WDR4 gene expression level, protein expression level and drug resistance phenotype in acute myeloid leukemia, as well as the direction and strength of the association.

7. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 6, characterized in that, Based on the results of drug resistance association analysis, the drug resistance status of patients with acute myeloid leukemia was determined, including: Key information was extracted from the drug resistance association analysis results, including the direction and strength of the association between WDR4 gene expression level and WDR4 protein expression level and drug resistance phenotype. At the same time, the critical values ​​for dividing the high-expression group and the low-expression group in the grouping analysis were confirmed, as well as the half-maximal inhibitory concentration (ICP-C) values ​​of the sensitive cell lines. Next, confirm the patient's core test data, including the relative expression level of the WDR4 gene, the relative expression level of the WDR4 protein, and the half-maximal inhibitory concentration (IC50) values ​​for different chemotherapy drugs. The patient's core test data were matched with the critical values ​​and thresholds determined in the drug resistance association analysis results; After matching, preliminary drug resistance data for the patients were obtained, including whether the patients' WDR4 gene expression level was high or low; whether the patients' WDR4 protein expression level was high or low; and whether the patients' half-maximal inhibitory concentration (IC50) values ​​for different chemotherapeutic drugs were higher than those of sensitive cell lines. Preliminary drug resistance data is used to determine a patient's sensitivity and resistance to a certain drug. Specifically, if a patient has high expression levels of both the WDR4 gene and WDR4 protein, and the half-maximal inhibitory concentration (IC50) is higher than the sensitivity threshold, the patient is considered to be in the high resistance range. If a patient has high expression levels of either the WDR4 gene or the WDR4 protein, and the IC50 is higher than the sensitivity threshold, the patient is considered to be in the intermediate resistance range, based on the correlation strength. If a patient has low expression levels of both the WDR4 gene and the WDR4 protein, and the IC50 is lower than the sensitivity threshold, the patient is considered to be in the sensitive range. The results of the preliminary drug resistance assessment will be verified by combining them with the patient's clinical information, which includes previous treatment response, relapse status, and disease progression. If the preliminary drug resistance data indicates high or low resistance, and the patient has previously experienced poor efficacy or relapse when using the drug, then the drug resistance should be reassessed. If the preliminary drug resistance data indicates a sensitive range, and the patient has previously responded well to the drug, then the drug resistance result should be confirmed. If the preliminary drug resistance data contradicts the clinical situation, then the reliability of the test data and correlation analysis should be re-examined. Finally, based on the combined validation results, the patient's drug resistance status was confirmed for each commonly used chemotherapy drug.

8. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 3, characterized in that, Using the synthesized cDNA product as a template, amplification was performed using WDR4 gene-specific primers and internal reference gene primers. During the amplification process, the kit monitored and recorded changes in fluorescence signal in real time and generated amplification curves. Based on the generated amplification curves, the Ct values ​​of the WDR4 gene and internal reference gene in each sample were automatically calculated. The Ct value is the cycle number at which the fluorescence signal reaches a set threshold, including: Using the synthesized cDNA product as a template, amplification was performed using WDR4 gene-specific primers and internal reference gene primers. During the amplification process, the acquisition window was dynamically adjusted based on the amplification stage to acquire fluorescence signals, and fluorescence acquisition signals at each amplification stage were obtained. Fluorescence acquisition signals at each amplification stage were analyzed using sliding window variance mutation detection to determine the fluctuation range of fluorescence signals. Based on the variance of the fluorescence signal fluctuation range, the Gaussian filter parameters are adaptively adjusted to reduce the noise of the fluorescence acquisition signals at each amplification stage, and the denoised fluorescence acquisition signals at each amplification stage are obtained. The fluorescence acquisition signals of each amplification stage after noise reduction are fitted to generate amplification curves; Based on the five-parameter dual asymptote model and the dynamic setting threshold of the amplification efficiency coefficient α, the basic Ct values ​​of the WDR4 gene and the internal reference gene in each sample were solved to obtain the basic Ct values ​​of the WDR4 gene and the internal reference gene in each sample. The first auxiliary Ct value and the second auxiliary Ct value of the WDR4 gene were obtained by extracting the local curvature maximum point of the ellipse fitting and the slope maximum point of the polynomial fitting from the WDR4 gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the WDR4 gene; The final Ct value of the WDR4 gene was obtained by fusion of the weights of the five-parameter double asymptote model error, the local error of ellipse fitting, and the polynomial fitting error. The first auxiliary Ct value and the second auxiliary Ct value of the internal reference gene are obtained by extracting the local curvature maximum point of ellipse fitting and the slope maximum point of polynomial fitting from the internal reference gene. Calculate the five-parameter double asymptote model error, ellipse fitting local error, and polynomial fitting error of the internal reference gene; The final Ct value of the internal reference gene is obtained by fusion of the weights adjusted in reverse based on the five-parameter dual asymptote model error, the local error of ellipse fitting, and the polynomial fitting error.

9. The method for detecting drug resistance in acute myeloid leukemia using WDR4 according to claim 4, characterized in that, Before using image analysis tools to analyze the grayscale values ​​of the WDR4 protein band and the internal reference protein band, the following steps are also included: screening the protein band images to obtain the target protein band image; The process of filtering protein band images to obtain target protein band images includes: The protein band image is denoised using a bilateral filtering algorithm to obtain a denoised protein band image; one of the denoised protein band images is randomly selected as the image to be processed. The image to be processed is divided into several independent swimlane images; Each individual swimlane image is converted to grayscale; The WDR4 region in each independent lane image is slid based on the first preset window, and peak detection is performed on each sliding window. When a peak appears in a first preset number of consecutive first windows, the first preset number of consecutive first windows are taken as the effective region of the WDR4 region. The size of the first preset window is larger than the size of the second preset window, and the first preset number is less than the second preset number. Based on the second preset window, the intrinsic parameter region in each independent lane image is slid, and peak detection is performed on each sliding window. When a second preset number of consecutive second windows all show peaks, the second preset number of consecutive second windows are taken as the effective region of the intrinsic parameter region. The distribution continuity of pixels in the effective regions of the WDR4 region and the internal reference region is determined respectively. If the number of effective regions in either region is greater than 1, the image quality of that lane is deemed unqualified and is not used as the target protein band image. For the determined swimlane image, the optical density corresponding to all pixels in the effective region of the WDR4 region and the effective region of the internal reference region are calculated respectively. Based on the optical density corresponding to all pixels in the effective region of the WDR4 region, the integrated optical density and signal-to-noise ratio of the effective region of the WDR4 region are determined. Based on the optical density corresponding to all pixels in the effective region of the internal reference region, the integrated optical density and signal-to-noise ratio of the effective region of the internal reference region are determined. The edge sharpness value of the effective region of the WDR4 region is calculated. When the signal-to-noise ratio of the WDR4 region of the image to be processed is greater than the first preset signal-to-noise ratio threshold, the signal-to-noise ratio of the internal reference region is greater than the second preset signal-to-noise ratio threshold, and the edge sharpness value of the WDR4 region is greater than the preset edge sharpness threshold, the image to be processed is taken as the target protein strip image.

10. A smart detection kit for WDR4 to detect drug resistance in acute myeloid leukemia, used in the method for detecting drug resistance in acute myeloid leukemia using WDR4 as described in claim 7, characterized in that the kit... include: The inner cavity (11) of the box has an inner cavity slot (12) at one end. A centrifuge (13) and a filter (14) are installed in the inner cavity (11). The centrifuge (13) automatically identifies the blood sample type and matches the centrifugation parameters through a sample weight sensor and AI algorithm. At the same time, it monitors the centrifugation speed and vibration frequency in real time. If an abnormality occurs, it will automatically stop and alarm through the control panel (17). The filter (14) integrates a laser particle size sensor to analyze the size of impurity particles in the filtrate in real time after filtration. If the purity does not meet the standard, it will automatically start secondary filtration or replace the filter membrane. At the same time, it will automatically adjust the filtration pressure according to the viscosity of the supernatant. The filtration progress is displayed on the control panel (17). The inner cavity slot (12) is equipped with a cooling rod (15) and a placement rod (16). The cooling rod (15) adopts semiconductor refrigeration technology and monitors the temperature in real time through a temperature sensor. It automatically adjusts the refrigeration power. At the same time, the cooling rod (15) is synchronized with the control panel (17) and automatically times according to the experimental steps. The outer wall of the kit is equipped with a control panel (17). The placement rod (16) has an optical sensor built in it to monitor the incubation status of the sample in the well plate in real time. The control panel (17) automatically collects the operating data of each device and generates a raw data report through the built-in algorithm.

Images