Apolipoprotein marker combination for evaluating individual radiosensitivity and application thereof

Abstract

The invention discloses an apolipoprotein marker combination for evaluating individual radiosensitivity and application thereof, and belongs to the technical field of biotechnology and nuclear radiation protection. Aiming at the technical problems that a genomics detection method and a functional experiment method for evaluating individual radiosensitivity have obvious limitations and the like at present, the invention screens and establishes a radiosensitivity prediction model based on serum apolipoprotein combinations (APOA1, APOA2 and APOA4) through a high-throughput mass spectrometry technology on the basis of a C57BL / 6J mouse half lethal dose radiation experiment for the first time; the clinical application value of the apolipoprotein combinations (APOA1, APOA2 and APOA4) as the radiosensitivity prediction marker is confirmed for the first time, and a novel molecular marker is provided for professional radiation protection monitoring and formulation of a precise tumor radiotherapy scheme.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of biotechnology and nuclear radiation protection, and particularly relates to an apolipoprotein marker combination for evaluating individual radiosensitivity and application thereof. BACKGROUND

[0002] In the contemporary society with accelerated development of nuclear energy and space exploration, nuclear radiation protection has become a core issue concerning national security and public health. The traditional radiation protection system is mainly based on group average dose limitation. However, the radiosensitivity of individuals differs significantly, and this limitation is particularly prominent in occupational exposure scenarios (such as nuclear industry workers) and extreme environments (such as space radiation). Epidemiological data shows that the cancer risk of occupational exposure groups or nuclear accident survivors can differ by 3-5 times under the same radiation dose. In critical scenarios such as nuclear emergency response and astronaut protection, this difference can lead to serious health consequences, including acute radiation sickness, increased long-term cancer risk, and organ function damage. The core value of individual radiosensitivity prediction lies in promoting the transformation of the protection mode from "group average" to "precise intervention". By pre-screening to identify high-sensitivity individuals, we can avoid their engagement in high-risk radioactive work and reduce radiation exposure risks from the source. Dynamic health monitoring of in-service personnel can also help to detect radiation damage signs in advance and take timely protective measures. At present, the National Aeronautics and Space Administration (NASA) has included radiosensitivity assessment in the astronaut selection criteria, and several countries have also integrated individual difference parameters into their nuclear emergency plans to improve the precision and effectiveness of radiation protection. It is worth noting that radiosensitivity assessment is not only applicable to occupational exposure groups, but also has important value for nuclear medicine patients. For example, in tumor radiotherapy, the response of patients to radiation differs significantly: some patients develop severe radiation damage (such as radiation pneumonitis and esophagitis) after receiving standard doses, while others have poor treatment efficacy due to radiation resistance. Therefore, individual radiosensitivity prediction is also of great significance for optimizing radiotherapy regimens, improving radiotherapy efficacy, and reducing radiotherapy side effects.

[0003] From the molecular mechanism, the difference in radiosensitivity is mainly due to the abnormality of biological processes such as DNA damage repair ability, oxidative stress defense system and cell cycle regulation network. As an important regulatory factor of these biological processes, apolipoprotein related to cholesterol metabolism pathway has been proved to be closely related to radiosensitivity. For example: APOBEC3G gene knockdown can enhance the radiosensitivity of various cancer cell lines, and its mechanism involves cell S phase arrest and G2 / M arrest and inhibition of DNA damage repair; ApoE knockout mice show persistent motor coordination disorders and cognitive impairment after radiation, and exogenous ApoE can significantly alleviate the neurotoxicity of ionizing radiation; In addition, intravenous infusion of apolipoprotein derived from bone marrow mesenchymal stem cells can rescue radiation-induced DNA damage and premature aging phenotype in mice, revealing the potential role of apolipoprotein in maintaining genome stability and delaying radiation-induced aging after radiation.

[0004] Current individual radiosensitivity prediction methods have obvious limitations, such as: genomic detection (such as ATM, TP53 mutation screening) can reveal genetic susceptibility, but the cost is high, the detection period is long, and only reflects static genetic information, which has the limitation of being unable to monitor dynamically; Functional experiments (such as clonogenic assay) can predict sensitivity by directly evaluating the proliferation ability of cells after radiation, but the cycle is longer and cannot reflect the complexity of the in vivo environment. In contrast, serum protein detection has the advantages of non-invasiveness, convenient detection (such as ELISA, antibody chip), and dynamic monitoring, which can reflect the physiological changes of the body after radiation exposure in real time, and is more suitable for clinical and occupational protection scenarios, becoming an ideal method for predicting individual radiosensitivity. High-density lipoprotein mediates reverse cholesterol transport through apolipoprotein (such as ApoA1), and its level is negatively correlated with the risk of cardiovascular disease, and has been proven to be useful for the auxiliary diagnosis of autoimmune diseases and cancer, suggesting the potential application value of apolipoprotein in disease diagnosis. A number of studies have shown that serum amyloid A (SAA), ApoE and other apolipoproteins are differentially expressed after X-ray or γ-ray exposure, and have been proven to be markers of radiation exposure. Although there is currently no report of serum apolipoprotein as an individual radiation sensitivity marker, these findings provide a theoretical basis for the development of serum apolipoprotein markers for individual radiosensitivity prediction. SUMMARY

[0005] In view of the fact that the current genomics detection methods and functional experimental methods for evaluating individual radiosensitivity have obvious limitations, and there is no report on serum apolipoprotein as an individual radiosensitivity marker, the present application first establishes a radiosensitivity prediction model based on serum apolipoprotein combination (APOA1, APOA2 and APOA4) by high-throughput mass spectrometry technology based on C57BL / 6J mouse half-lethal dose radiation experiment. The expression profile between the radiosensitive group (dead mice) and the radiation-resistant group (surviving mice) is successfully identified by differential proteomics analysis. The prediction model constructed based on the XGBoost machine learning algorithm and combined with ROC curve analysis shows excellent discrimination ability in the training set, and the prediction performance is confirmed in another independent verification queue. The present application first confirms the clinical application value of the apolipoprotein combination (APOA1, APOA2 and APOA4) as a radiosensitivity prediction marker, and provides a new type of molecular marker for occupational radiation protection monitoring and formulation of precise radiotherapy plan.

[0006] To solve the above technical problems and achieve the corresponding technical effects, the present application provides the following technical solutions: The first object of the present application is to provide an apolipoprotein marker combination for evaluating individual radiosensitivity, which consists of APOA1, APOA2 and APOA4.

[0007] The second object of the present application is to provide the use of the above-mentioned apolipoprotein marker combination in the preparation of a product for evaluating individual radiosensitivity.

[0008] In an embodiment of the present application, the evaluation also requires collecting a sample from the evaluated person, and the sample is serum.

[0009] The third object of the present application is to provide a reagent for detecting the expression level of the above-mentioned apolipoprotein marker combination.

[0010] In an embodiment of the present application, the reagent is used to detect the expression level of each protein in the apolipoprotein marker combination.

[0011] In an embodiment of the present application, the expression level is the protein expression level and / or mRNA transcription level.

[0012] The fourth object of the present application is to provide the use of the above-mentioned reagent in the preparation of a kit for evaluating individual radiosensitivity.

[0013] In an embodiment of the present application, the evaluation also requires collecting a sample from the evaluated person, and the sample is serum.

[0014] A fifth object of the present invention is to provide a kit containing the above-described reagents for assessing an individual's radiosensitivity.

[0015] In one embodiment of the invention, the assessment also requires the collection of a sample from the person being assessed, the sample being serum.

[0016] The beneficial effects of this invention are: This invention, for the first time, utilizes high-throughput mass spectrometry to screen and establish a radiosensitivity prediction model based on a combination of serum apolipoproteins (APOA1, APOA2, and APOA4) in a C57BL / 6J mouse median lethal dose radiation experiment. Differential proteomics analysis successfully identified the expression profiles between the radiosensitive group (deceased mice) and the radioresistant group (surviving mice). The prediction model, constructed using the XGBoost machine learning algorithm combined with ROC curve analysis, demonstrated excellent discriminative ability in the training set, and this predictive efficacy was confirmed in another independent validation cohort. This invention is the first to demonstrate the clinical application value of apolipoprotein combination (APOA1, APOA2, and APOA4) as a radiosensitivity prediction biomarker, providing novel molecular markers for occupational radiation protection monitoring and the development of precise radiotherapy protocols for tumors. Attached Figure Description

[0017] Figure 1 The image shows the 30-day Kaplan-Meier survival curves of mice irradiated at the median lethal dose in the training and validation cohorts; among them, Figure 1 In the figure, A represents the 30-day Kaplan-Meier survival curve of mice in the training cohort subjected to the median lethal dose of irradiation. Figure 1 B in the figure represents the 30-day Kaplan-Meier survival curve of mice in the validation cohort subjected to the median lethal dose of irradiation. Figure 2 Box plots showing the initial body weight of mice in the training and validation cohorts before ionizing radiation exposure; where, Figure 2 In the figure, A is a box plot of the initial body weight of mice in the training cohort before ionizing radiation, representing the median lethal dose irradiation model. Figure 2 In the diagram, B represents the initial body weight of mice in the validation cohort before ionizing radiation; SV: survival resistant group, Death: death resistant group; Figure 3 The graph shows the results of the correlation analysis between initial body weight before ionizing radiation and survival time after irradiation in mice of the median lethal dose irradiation model in the training and validation cohorts; among them, Figure 3 Figure A in the graph represents the correlation between initial body weight before ionizing radiation and survival time after irradiation in mice of the median lethal dose irradiation model in the training cohort. Figure 3Figure B in the figure represents the results of the analysis of the correlation between initial body weight before ionizing radiation and survival time after irradiation in mice of the median lethal dose irradiation model in the validation cohort. Figure 4 A bar chart showing the top 20 KEGG pathway enrichment entries for differentially expressed proteins in the serum of mice in the death-sensitive and survival-resistant groups of the training cohort; where the X-axis represents the number of differentially expressed proteins, the Y-axis represents the KEGG pathway name, and different colors in the figure correspond to different pathway categories: red represents Metabolism, light blue represents Genetic Information Processing, green represents Environmental Information Processing, dark blue represents Cellular Processes, and pink represents Organismal Systems. Figure 5 To validate the KEGG pathway enrichment of the top 20 differentially expressed proteins in the serum of mice in the death-sensitive and survival-resistant groups in the cohort, a bar chart was created. The X-axis represents the number of differentially expressed proteins, and the Y-axis represents the KEGG pathway name. Different colors in the chart correspond to different pathway categories: red represents Metabolism, light blue represents Genetic Information Processing, green represents Environmental Information Processing, dark blue represents Cellular Processes, and pink represents Organismal Systems. Figure 6 A bubble plot of the first 20 GO enrichment biological processes of differentially expressed proteins in the serum of mice in the death-sensitive and survival-resistant groups in the training cohort; where the X-axis is -log10(p) (the negative logarithm of the P-value, the larger the value, the more significant the enrichment), the Y-axis is the GO biological process item, the bubble size represents the number of differentially expressed proteins (Count), the larger the bubble, the more differentially expressed proteins involved in the biological process, and the color gradient represents the Fold Enrichment, the darker the color, the higher the fold enrichment. Figure 7To validate the GO enrichment biological process of differentially expressed proteins in the serum of mice in the death-sensitive and survival-resistant groups in the cohort, a bubble plot of the first 20 entries was created. The X-axis represents -log10(p) (the negative logarithm of the P-value, with a larger value indicating more significant enrichment), the Y-axis represents the GO biological process entries, the bubble size represents the number of differentially expressed proteins (Count), with larger bubbles indicating a greater number of differentially expressed proteins involved in the biological process, and the color gradient represents the Fold Enrichment (fold increase), with darker colors indicating a higher fold increase. Figure 8 The graph shows the ROC curves for predicting mouse radiosensitivity in the training cohort using the XGBoost algorithm, with apolipoproteins Apoa1, Apoa2, and Apoa4 individually and in combination. Figure 8 In the figure, A represents the ROC curve of predicting mouse radiosensitivity using the apolipoprotein Apoa1 alone in the training cohort based on the XGBoost algorithm. Figure 8 In the figure, B represents the ROC curve of predicting mouse radiosensitivity using apolipoprotein Apoa2 alone in the training cohort based on the XGBoost algorithm. Figure 8 C in the figure represents the ROC curve of predicting mouse radiosensitivity using apolipoprotein Apoa4 alone in the training cohort based on the XGBoost algorithm. Figure 8 D in the figure represents the ROC curve of the combination of apolipoproteins Apoa1, Apoa2 and Apoa4 in the training cohort based on the XGBoost algorithm to predict the radiosensitivity of mice. The area under the curve (AUC) and its 95% confidence interval (CI) are marked in the figure to measure the predictive performance of the model. The closer the AUC value is to 1, the stronger the model's discriminative ability. Figure 9 ROC curves for predicting mouse radiosensitivity in a validation cohort using the XGBoost algorithm with apolipoproteins Apoa1, Apoa2, and Apoa4 individually and in combination. Figure 9 In the figure, A represents the ROC curve of predicting mouse radiosensitivity using apolipoprotein Apoa1 alone in the validation cohort based on the XGBoost algorithm. Figure 9 In the graph, B represents the ROC curve of predicting mouse radiosensitivity using apolipoprotein Apoa2 alone in the validation cohort based on the XGBoost algorithm. Figure 9 C in the figure represents the ROC curve of predicting mouse radiosensitivity using apolipoprotein Apoa4 alone in the validation cohort based on the XGBoost algorithm. Figure 9D in the figure represents the ROC curve of predicting the radiosensitivity of mice in the validation cohort using the combination of apolipoproteins Apoa1, Apoa2, and Apoa4 based on the XGBoost algorithm. The area under the curve (AUC) and its 95% confidence interval (CI) are marked in the figure to measure the predictive performance of the model. The closer the AUC value is to 1, the stronger the model's discriminative ability. Figure 10 This is a box plot showing the relative expression levels of each protein (APOA1, APOA2, APOA4) in a serum apolipoprotein biomarker combination based on an individual radiosensitivity prediction model in the survival-resistant and death-sensitive mouse populations of the training and validation cohorts; among them, Figure 10 In the figure, A represents a box plot showing the relative expression levels of serum apolipoprotein APOA1 in the survival-resistant and death-sensitive mouse populations of the training cohort. Figure 10 B in the figure represents a box plot showing the relative expression levels of serum apolipoprotein APOA2 in the survival-resistant and death-sensitive mouse populations of the training cohort. Figure 10 In the figure, C represents the box plot of the relative expression levels of serum apolipoprotein APOA4 in the survival-resistant and death-sensitive mouse populations of the training cohort. Figure 10 In the figure, D represents the box plot of the relative expression level of serum apolipoprotein APOA1 in the survival-resistant and death-sensitive mouse populations of the validation cohort. Figure 10 E in the figure represents a box plot showing the relative expression levels of serum apolipoprotein APOA2 in the survival-resistant and death-sensitive mouse populations of the validation cohort. Figure 10 F in the figure represents a box plot of the relative expression level of serum apolipoprotein APOA4 in the survival-resistant and death-sensitive mouse populations of the validation cohort. In each subplot, "SV" on the horizontal axis represents the survival-resistant group and "D" represents the death-sensitive group, and the vertical axis represents the relative expression level. The values ​​marked above the subplots are the p-values ​​for comparing the differences between the two groups, used to illustrate the statistical significance of the differences. The box plots show the median and quartile information of the data, and the scatter plots represent the specific test values ​​of each sample. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings. It should be noted that the embodiments mentioned below are only for explaining the invention and are not intended to limit the scope of the invention. The embodiments mentioned below are only some embodiments of the invention, not all embodiments. Those skilled in the art can refer to the content of this document and appropriately improve the process parameters to achieve the objectives of the invention. It should be particularly noted that all similar substitutions and modifications are obvious to those skilled in the art and are considered to be included in the invention. The methods and applications of this invention have been described through preferred embodiments, and those skilled in the art can obviously modify or appropriately change and combine the methods and applications described herein without departing from the content and scope of this invention to realize and apply the technology of this invention. In the art, embodiments obtained by other those skilled in the art without creative effort are all protected by this invention.

[0019] Unless otherwise specified, the experimental methods used in the following examples are conventional methods, and the materials, reagents and instruments used are conventional materials, reagents and instruments in the art, which can be obtained by those skilled in the art through commercial channels.

[0020] Example 1: Screening and validation of biomarker combinations for individual radiosensitivity assessment 1. Establishment of a mouse median lethal dose irradiation model This invention uses SPF-grade male C57BL / 6J mice (8-10 weeks old, weighing 22-25 g) purchased from Spefol (Beijing) Biotechnology Co., Ltd. as experimental subjects, and conducts two batches of 7.3 Gy 60 A half-lethal dose irradiation model was established by whole-body irradiation with Co γ rays (training cohort n=60, 61.54 cGy / min; validation cohort n=24, 60.75 cGy / min).

[0021] 2. Radiation susceptibility analysis of mice with median lethal dose irradiation model All mice underwent baseline blood collection 7 days prior to irradiation (100 μL of blood was collected from the tail vein, and 30 μL of serum was obtained by centrifugation). Each independent cohort was divided into a radiation-sensitive death group and a radiation-resistant survival group based on the survival outcome 30 days after irradiation. To assess the radiation sensitivity of C57BL / 6J mice, survival analysis was performed after irradiation. Daily survival status was recorded for 30 days after irradiation (death events were defined as experimental endpoints). Kaplan-Meier survival curves were constructed to reflect the effect of radiation on mouse survival and the degree of radiation tolerance in mice: (1) mortality rate over time (represented by a stepwise decrease); (2) cumulative mortality rate (represented by the vertical decrease); (3) radiation-resistant group (represented by the formation of a plateau). To exclude the influence of body weight on the outcome of mice after irradiation, the body weight on the day of irradiation was recorded simultaneously in the experiment, and Pearson correlation analysis was used to verify the independence of initial body weight and survival time.

[0022] To assess individual differences in radiosensitivity, this invention uses 7.3 Gy 60 A median-lethal dose (LD50) irradiation model was established in C57BL / 6J mice by whole-body Co-γ irradiation. Two independent experimental groups were established: a training cohort (n=60) and a validation cohort (n=24). The 30-day survival rate in the training cohort was 36.7% (22 mice survived, 38 died), while the 30-day survival rate in the validation cohort was 62.5% (15 mice survived, 9 died). Figure 1 There were no statistically significant differences in initial body weight between the survival resistance group (SV) and the death-sensitive group in both cohorts (body weight of the survival resistance group in the training cohort was 24.1±1.1 g, and body weight of the death-sensitive group was 24.5±0.9 g, P=0.155; body weight of the survival resistance group in the validation cohort was 23.5±0.8 g, and body weight of the death-sensitive group was 24.0±0.3 g, P=0.078). Figure 2 This effectively ruled out the potential influence of body weight on irradiation outcomes. Further Pearson correlation analysis was used to verify the correlation between initial body weight and post-irradiation survival time. The results showed that there was a weak and insignificant correlation between mouse body weight and post-irradiation survival time in both cohorts (correlation coefficient r = -0.189, p = 0.149 in the training cohort; correlation coefficient r = -0.355, p = 0.089 in the validation cohort), indicating that body weight had no effect on irradiation survival time. Figure 3 Although the final survival rates differed between the two batches of experiments, which could be attributed to slight fluctuations in dose rate and changes in environmental parameters (temperature, humidity), both models generally exhibited typical median lethal dose characteristics, confirming that 7.3 Gy 60Co-γ irradiation can serve as a stable model for evaluating radiosensitivity. Serum samples from the training cohort are used for differential protein screening and machine learning model construction, while serum samples from the validation cohort are used for independent validation of key protein biomarkers, thereby ensuring the reliability of the research results.

[0023] 3. Mass spectrometry sequencing Serum proteomics analysis was performed using a Thermo Scientific Orbitrap Astral mass spectrometer coupled with a nanoLC 1200 system for data acquisition-independently (DIA). Samples underwent denaturation, reduction, and alkylation treatments, followed by trypsin digestion. The resulting peptides were desalted, eluted with an acetonitrile gradient, separated by nano-liquid chromatography, and then analyzed on the Orbitrap Astral mass spectrometer in DIA mode. Raw data were processed using DIA-NN1.8 software, and the UniProtKB Mus musculus database was searched. The false discovery rate (FDR) for peptide mapping (PSM) and protein levels was set to 1%. Differentially expressed proteins were selected based on a fold change ≥ 1.5 and a Student's staunch test p < 0.05.

[0024] 4. Functional enrichment analysis By analyzing the KEGG pathway and GO biological processes of differentially expressed serum proteins in mice from the death-sensitive and survival-resistant groups... Enrichment analysis was performed to reveal the potential association between differentially expressed proteins related to metabolic processes and radiosensitivity.

[0025] The Wilcoxon rank-sum test was used to examine the differences in serum protein expression between the mortality-sensitive group and the survival-resistant group. Enrichment analysis was performed on proteins with P < 0.05 and |log2FC| ≥ 0.585 using the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology Biological Processes (GO). KEGG pathway enrichment analysis was performed using the clusterProfiler software package, and pathway significance was assessed using the hypergeometric test. After Benjamini-Hochberg correction, P < 0.05 was used as the significance threshold. GO biological process analysis was also performed using the clusterProfiler tool, and the same statistical methods were used to screen for significantly enriched items (P < 0.05). The results were visualized using ggplot2, showing the top 20 most significantly enriched pathways or biological processes.

[0026] To screen for differentially expressed proteins related to radiosensitivity, we performed KEGG and GO enrichment analyses on baseline differentially expressed proteins (P<0.05, |log2FC|≥0.585) between the death-sensitive and survival-resistant groups in both the training and validation cohorts. KEGG pathway enrichment analysis showed that the metabolic pathway was significantly enriched in both the training and validation cohorts. Figure 4 and Figure 5 GO enrichment analysis showed that metabolic-related biological processes (such as cellular metabolism and organic nitrogen compound metabolism) were significantly enriched in both the training and validation cohorts, with a large number of differentially expressed proteins and high fold enrichment. Figure 6 and Figure 7 This suggests that differentially expressed proteins related to metabolic processes may be closely related to radiosensitivity.

[0027] 5. Machine learning predicts key proteins of radiosensitivity Machine learning analysis employed the XGBoost algorithm to screen key predictive proteins for radiosensitivity. Protein expression data obtained from mass spectrometry quantification in the training cohort was used as the training set, and protein expression data obtained from mass spectrometry quantification in the validation cohort was used as the validation set. All data underwent Z-score normalization. Model construction utilized the xgboost package (v1.7.7) in R (4.4.2), with the following basic parameters set: learning rate (eta) = 0.1, maximum tree depth (max_depth) = 6, subsample ratio (subsample) = 1, minimum leaf node sample size (min_child_weight) = 1, and number of iterations (nrounds) = 100. Model performance was evaluated using the area under the receiver operating characteristic (AUC), and 95% confidence intervals were calculated using 2000 Bootstrap resampling iterations. Differentially expressed proteins were selected based on their XGBoost feature importance score (Gain value) ranking in the top 10% and a permutation test p < 0.01.

[0028] 1) A robust XGBoost model based on a combination of metabolically related serum apolipoproteins Apoa1 (UniprotKB: Q00623), Apoa2 (UniprotKB: P09813), and Apoa4 (UniprotKB: Q810I7) was used to predict radiosensitivity in mice. (UniProt website) https: / / www.uniprot.org / (Provided login ID) The above findings confirm that metabolic process-related proteins may be associated with radiosensitivity. To establish a predictive model for mouse radiosensitivity, we trained a machine learning model using the XGBoost algorithm based on differentially expressed metabolic process proteins between the mortality-sensitive group and the survival-resistant group in the training cohort (P < 0.05 and |log2FC| ≥ 0.585). In the training cohort, the area under the receiver operating characteristic (ROC) curves (AUC) for apolipoproteins Apoa1, Apoa2, and Apoa4 individually predicting mouse radiosensitivity were 0.870 (95% CI: 0.745–0.994), 0.880 (95% CI: 0.764–0.997), and 0.824 (95% CI: 0.712–0.936), respectively. Figure 8 A in Figure 8 B in Figure 8 These results indicate that all three apolipoproteins possess some predictive ability, with Apoa1 and Apoa2 showing relatively better predictive performance. After using the XGBoost algorithm to perform machine learning modeling on the combination of the three apolipoproteins, the resulting ROC curve had an AUC value of 1 (95% CI: 1-1) in the training queue. Figure 8 The D in the figure indicates that the combination has extremely strong predictive and discriminative ability. In the validation cohort, the AUCs of the receiver operating characteristic (ROC) curves of apolipoproteins Apoa1, Apoa2, and Apoa4 individually predicting mouse radiosensitivity were 0.904 (95% CI: 0.783–1), 0.926 (95% CI: 0.803–1), and 0.704 (95% CI: 0.487–0.921), respectively. Figure 9 A in Figure 9 B in Figure 9 In the C group, Apoa1 and Apoa2 maintained good predictive performance. The XGBoost model based on the combination of three apolipoproteins continued to perform well in the validation cohort, with an AUC value of 0.926 (95% CI: 0.826–1.000) for its ROC curve. Figure 9 The D in the figure indicates that the model has good predictive discrimination ability in independent validation. In summary, the results show that the XGBoost model based on the combination of Apoa1, Apoa2, and Apoa4 exhibits good predictive performance for mouse radiosensitivity in both the training and validation cohorts. Notably, Apoa1 and Apoa2 have shown stable predictive performance as single markers, which may suggest that Apoa1 and Apoa2 play a core role in this predictive combination.

[0029] 2) Differential expression characterization of serum apolipoprotein combinations (APOA1, APOA2, APOA4) in radiosensitivity prediction By analyzing the expression characteristics of serum apolipoprotein combinations (APOA1, APOA2, APOA4) in an individual radiosensitivity prediction model, the results showed that in the training cohort, the expression levels of APOA1 (P<0.0001) and APOA2 (P<0.0001) in the survival resistance group were significantly higher than those in the death group. Figure 10 A in Figure 10 The same trend was observed in the verification queue (APOA1: P=0.0002; APOA2: P=0.0002). Figure 10 D and Figure 10 The expression of APOA1 and APOA2 may confer radiation resistance in mice, suggesting that high expression of APOA1 and APOA2 may confer radiation resistance. Conversely, the expression pattern of APOA4 in the training cohort (P<0.0001) and validation cohort (P=0.0224) was the opposite, with lower expression levels in the survival group than in the death group. Figure 10 C and Figure 10 The F in the figure suggests that high expression of APOA4 is associated with radiosensitivity in mice. Notably, both APOA1 and APOA2 showed significant inter-group differences in the training and validation cohorts, with low overlap in their relative expression distributions. This is highly consistent with the predictions of the XGBoost model, further validating their potential as biomarkers of radiosensitivity.

[0030] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A combination of apolipoprotein biomarkers for assessing individual radiosensitivity, characterized in that, The apolipoprotein biomarker group consists of APOA1, APOA2 and APOA4.

2. The use of the apolipoprotein biomarker combination of claim 1 in the preparation of products for assessing individual radiosensitivity.

3. The application according to claim 2, characterized in that, The assessment also requires the collection of a sample from the person being assessed, which is serum.

4. A reagent for detecting the expression level of the combination of apolipoprotein biomarkers of claim 1.

5. The reagent according to claim 4, characterized in that, The reagents are used to detect the expression levels of each protein in the apolipoprotein biomarker combination.

6. The reagent according to claim 5, characterized in that, The expression level refers to the protein expression level and / or mRNA transcription level.

7. Use of the reagent according to any one of claims 4-6 in the preparation of a kit for assessing individual radiosensitivity.

8. The application according to claim 7, characterized in that, The assessment also requires the collection of a sample from the person being assessed, which is serum.

9. A kit for assessing individual radiosensitivity containing the reagents of any one of claims 4-6.

10. The reagent kit according to claim 9, characterized in that, The assessment also requires the collection of a sample from the person being assessed, which is serum.

Images