Screening method and application of oligoasthenospermia biomarker
The biomarkers of oligospermia were screened through metabolomics and machine learning methods, and a logistic regression model was constructed, which solved the problem of difficulty in diagnosing oligospermia in the existing technology, achieving efficient screening and diagnostic effects.
Patent Information
- Application Number
- CN202411936808.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-26
- Publication Date
- 2025-04-08
AI Technical Summary
The prior art is difficult to effectively screen and diagnose biomarkers of oligospermia, and lacks efficient diagnostic methods.
The semen pulp samples were analyzed by non-targeted metabolomics, and the candidate biomarkers 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol were screened out, and a logistic regression model was constructed in combination with an integrated machine learning method to distinguish oligospermia from normal sperm.
Efficient screening and diagnosis of oligospermia is achieved, and the sensitivity, specificity and accuracy of the model are 0.93, 1.0 and 0.97 respectively, providing a reliable diagnostic tool.
Smart Images
Figure CN119785890A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of medicine and molecular diagnosis technology, and in particular relates to a method for screening a biomarker of oligoasthenozoospermia and an application thereof. Background Art
[0002] In the field of male infertility, semen is the most commonly used biological fluid for the assessment of male fertility. Several excellent biomarkers associated with male infertility have been established by evaluating semen, such as high sperm DNA fragmentation index and lipid peroxidation. Metabolomic analysis of seminal plasma for the prediction of male fertility disorders has also been shown to be a powerful technique as it provides relevant information for the diagnosis of male factor infertility.
[0003] In order to further study the molecular mechanism of low sperm motility in patients with oligoasthenozoospermia and discover biomarkers for diagnosis of oligoasthenozoospermia and potential targets for treatment, the present invention applies a metabolomics method to further study patients with oligoasthenozoospermia and provides a screening method for oligoasthenozoospermia biomarkers and its application. Summary of the invention
[0004] The purpose of the embodiments of the present invention is to provide a method for screening biomarkers of oligoasthenozoospermia and its application, aiming to solve the problems raised in the above-mentioned background technology.
[0005] The present invention implements a method for screening biomarkers of oligoasthenozoospermia through the following steps: Step 1: Perform metabolomics analysis of seminal plasma in oligoasthenozoospermia patients and normal control group: Seminal plasma samples from oligoasthenospermia patients and normal men were collected for non-targeted metabolomics testing. According to the significance criteria of multidimensional statistical analysis and univariate statistical analysis, i.e., VIP>1, p When value is less than 0.05, the significant difference of metabolites between the two groups was statistically analyzed to screen candidate biomarkers; Step 2: Biomarker screening for oligoasthenozoospermia: By integrating machine learning methods, the comprehensive weight value of each metabolite was calculated, and the candidate biomarkers were ranked according to their comprehensive weight values. The receiver operating characteristic (ROC) analysis was used to evaluate the influence of each metabolite on the model AUC value. The AUC cumulative curve of the candidate biomarkers showed that the cumulative AUC values of the top two metabolites were the largest. Therefore, the first two metabolites, 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol, formed the best candidate biomarker combination, which may be a new potential biomarker group for diagnosing oligoasthenospermia.
[0006] Step 3: Verify and characterize candidate biomarkers: Logistic regression (LR), random forest (RF) and support vector machine (SVM) were used to verify the above-screened marker combinations; Step 4: Build a diagnostic model through logistic regression to distinguish oligoasthenozoospermic sperm from normal sperm, and evaluate the diagnostic ability of the diagnostic model: The optimal cutoff calculated by the maximum Youden index was 0.68, corresponding to the sensitivity, specificity and accuracy of the diagnostic model of 0.93, 1.0 and 0.97, respectively. 70% of the original cohort was used as a training set to train the model, and then the trained model was used for the remaining 30% of the prediction. ROC analysis was performed using the diagnostic model constructed above based on the expression levels of candidate biomarkers in the samples.
[0007] In a further technical solution, in step 4, the biomarker logistic regression model formula constructed is: ; in, z The expression is as follows: ; Substitute the expression level values of the two biomarkers in the sample to be classified into the above model formula to calculate the probability value P ,if P If the value exceeds the cutoff value, it is classified as the oligoasthenozoospermia group, otherwise it is classified as the normal group.
[0008] Another objective of an embodiment of the present invention is to propose an application of a biomarker for oligoasthenozoospermia, based on the biomarker combination screened by the above method, namely 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol, which is used to prepare a kit and a detection device for identifying male oligoasthenozoospermia.
[0009] The embodiments of the present invention provide a method for screening a biomarker for oligoasthenozoospermia and its application, which uses non-targeted metabolomics to systematically analyze seminal plasma metabolites. The method identified a large number of metabolites and metabolic pathways related to oligoasthenozoospermia, providing a comprehensive perspective on the changes in metabolites of patients with oligoasthenozoospermia. A group of biomarker combinations that can be used to identify oligoasthenozoospermia were screened out, and they were used to prepare kits and detection equipment for diagnosing oligoasthenozoospermia, providing an effective auxiliary means for the diagnosis of oligoasthenozoospermia. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Figure 1 The present invention provides a flowchart of a method for screening a biomarker for oligoasthenozoospermia in an embodiment of the present invention.
[0011] Figure 2 In the positive / negative ion mode, the OPLS-DA score graphs of the two groups of seminal plasma samples and their permutation test results are shown respectively; Figure 3 This is the hierarchical clustering heat map of significantly different metabolites in positive / negative ion mode; Figure 4 Rank the weight values of candidate biomarkers; Figure 5 It is the AUC cumulative trend graph of the candidate Biomarkers classification model; Figure 6 Results of AUC, specificity, and sensitivity of the three single models are shown; Figure 7 Random forest calculates the importance of Biomarkers; Figure 8 are the expression levels of the two biomarkers in the oligoasthenozoospermia group and the normal control group; Fig. 9 Correlation analysis diagram of candidate Biomarkers; Fig.10 Trend graph between model accuracy, specificity, sensitivity and cutoff; Fig.11 It is the ROC curve diagram. DETAILED DESCRIPTION
[0012] In order to more clearly understand the purpose, technical solutions and advantages of the present invention, the present invention will be described in more detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to illustrate the content of the present invention and do not limit the scope of the present invention.
[0013] Next, the implementation process of the present invention will be described in detail in conjunction with specific embodiments.
[0014] like Figure 1 As shown, a method for screening a biomarker for oligoasthenozoospermia proposed in one embodiment of the present invention is shown, and the method comprises the following steps: Step 1: Perform metabolomics analysis on seminal plasma samples from the oligoasthenozoospermia group and the normal control group.
[0015] Seminal plasma samples from oligoasthenospermia patients and normal men were collected for non-targeted metabolomics testing. According to the significance criteria of multivariate statistical analysis and univariate statistical analysis (VIP>1, p value<0.05), the metabolite differences between the two groups were statistically analyzed to screen out candidate biomarkers.
[0016] Step 2: Biomarker screening for oligoasthenozoospermia: By integrating machine learning methods, the comprehensive weight value of each metabolite was calculated, and the candidate biomarkers were ranked according to their comprehensive weight values. Receiver operating characteristic analysis was used to evaluate the influence of each metabolite on the AUC value of the model. The cumulative AUC curve of the candidate biomarkers showed that the cumulative AUC values of the top two metabolites were the largest. Therefore, the first two metabolites, 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol, formed the best candidate biomarker combination, which may be a new potential biomarker group for diagnosing oligoasthenozoospermia.
[0017] Step 3: Verify and characterize candidate biomarkers: In order to evaluate the effect of the above two candidate biomarker combinations on the classification model, logistic regression, random forest and support vector machine were used to verify the above selected marker combinations. The sensitivity, specificity and accuracy of the three differential models were not less than 0.95, which indicated that these two metabolites could be used as effective biomarkers for screening oligoasthenozoospermia.
[0018] Step 4: Build a diagnostic model through logistic regression to distinguish oligoasthenozoospermic sperm from normal sperm, and evaluate the diagnostic ability of the diagnostic model: The optimal cutoff calculated by the maximum Youden index was 0.68, corresponding to the sensitivity, specificity and accuracy of the diagnostic model of 0.93, 1.0 and 0.97, respectively. 70% of the original cohort was used as a training set to train the model, and then the trained model was used for the remaining 30% of the prediction. ROC analysis was performed using the diagnostic model constructed above based on the expression levels of candidate biomarkers in the samples.
[0019] As a preferred embodiment of the present invention, in step 1, semen of 30 oligoasthenozoospermia patients and 30 normal control males were collected, and non-targeted metabolomics analysis was performed using ultra-high performance liquid chromatography-tandem time-of-flight mass spectrometry (UHPLC-Q-TOFMS). After combining positive and negative ion modes, a total of 2201 metabolites were identified, of which 1331 metabolites were identified in the positive ion mode and 870 metabolites were identified in the negative ion mode.
[0020] Univariate statistical analysis was performed to analyze the differences of all metabolites (including unidentified metabolites) detected in the seminal plasma samples of oligoasthenozoospermia patients and normal control men in positive and negative ion modes. Among the significantly different metabolites, those with FC > 1.5 in the oligoasthenozoospermia group compared with the normal control group were p value < 0.05 There are 128 different metabolites, which meet the FC < 0.67 between the oligoasthenozoospermia group and the normal control group. p There were 83 differential metabolites with value < 0.05. Through orthogonal partial least squares discriminant analysis (OPLS-DA), the maximum difference between groups was reflected in the t[1] dimension, while the difference within the group was reflected in the to[1] dimension. Figure 2 As shown in A and 2B, the seminal plasma metabolite samples of oligoasthenozoospermia patients and normal control men showed obvious classification differences in both positive and negative ion modes. In order to avoid overfitting of the supervised model during the modeling process, the permutation test was used to verify the model to ensure its effectiveness. Figure 2 C and 2D show the permutation test results of the two OPLS-DA models. As the number of permutations increases, the R² and Q² values of the random model gradually decrease, indicating that the original model does not have overfitting and the model is robust.
[0021] The variable weight value (VIP) in the OPLS-DA model is used to measure the contribution and explanatory power of each metabolite in sample classification, so as to identify biologically significant differential metabolites. Generally, when the VIP value is greater than 1, the metabolite is considered to have a significant contribution to the model. In metabolomics analysis, VIP>1 andp The value < 0.05 was used as the standard for screening significantly different metabolites. According to this standard, 211 significantly different metabolites were screened out from the comparison between the oligoasthenozoospermia group and the normal control group.
[0022] Significantly different metabolites (VIP>1, p The results of hierarchical clustering analysis with value < 0.05 are as follows Figure 3 shown.
[0023] As a preferred embodiment of the present invention, in step 2, an integrated machine learning method is used to reward metabolites that are frequently selected in the feature selection method, and the comprehensive weight value of each metabolite is calculated by combining the reward scores obtained in each feature selection method. The larger the weight value, the greater the contribution of the substance in distinguishing the oligoasthenozoospermia group from the normal control group.
[0024] The ranking of the comprehensive weight values of candidate biomarkers in the seminal plasma of men with oligoasthenozoospermia is as follows Figure 4 As shown. The receiver operating characteristic (ROC) analysis was used to evaluate the effect of each metabolite on the area under the curve (AUC). The higher the AUC value, the better the classification effect of the model. Figure 5 As shown in the figure, the cumulative AUC curve of the candidate biomarkers shows that the cumulative AUC values of the top two metabolites are the largest. This indicates that the top two metabolites (1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol) can be used as candidate biomarkers of seminal plasma metabolites in patients with oligoasthenospermia.
[0025] As a preferred embodiment of the present invention, in step 3, in order to evaluate the effect of the above two candidate biomarker combinations on the classification model, three commonly used machine learning models, including logistic regression (LR), random forest (RF) and support vector machine (SVM), are used to verify the above screened biomarker groups. ROC curve analysis is used to determine the classification performance of the above two candidate biomarker combinations for male samples of oligoasthenozoospermia group and normal control group. Figure 6The results of AUC, specificity and sensitivity of the three models are shown. The evaluation indicators of AUC, specificity and sensitivity of the three single machine learning models are all ideal, indicating that the two biomarker combinations selected have relatively excellent classification ability and effect.
[0026] The classification model constructed by the random forest algorithm was used to calculate the importance coefficient of candidate Biomarkers, which was used to compare the contribution of each Biomarker to the model. Figure 7 ,The higher importance coefficient of the candidate Biomarkers indicates that the relationship between the Biomarker and the classification group is closer, and the greater contribution to distinguishing different groups.
[0027] The expression levels of the two biomarkers in the oligoasthenozoospermia group and the normal control group, such as Figure 8 As shown in the figure, it can be observed that the two biomarkers screened showed significant differential expression in the two groups of samples. In the oligoasthenozoospermia group, 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine was decreased while [6]-gingerol was increased.
[0028] Next, the Pearson correlation coefficient was calculated for the expression levels of the two biomarkers mentioned above. The correlation between the two biomarkers was low, which indicates that the overlap between the selected biomarkers is low and the combination of the panel is more optimized, such as Fig. 9 shown.
[0029] As a preferred embodiment of the present invention, in step 4, a diagnostic model of biomarkers is constructed using a logistic regression algorithm. Table 1 shows the logistic regression coefficients of the biomarkers.
[0030] Table 1 Logistic regression coefficients of biomarkers ; The biomarker logistic regression model formula is: ; in, z The expression is as follows: ; Substitute the expression level values of the two biomarkers in the sample to be classified into the above model formula to calculate the probability value P ,if PIf the value exceeds the cutoff value, it is classified as oligoasthenozoospermia group, otherwise it is classified as normal group. In order to obtain this boundary value, we use Youden's index to define the optimal cutoff value for diagnosis. The trend relationship between the accuracy, specificity, sensitivity and cutoff of the model is as follows Fig.10 Here, the cutoff value when the value of "sensitivity + specificity - 1" is the largest is used as the optimal cutoff value. The cutoff value finally defined in this analysis is 0.68, and its corresponding indicators are shown in Table 2.
[0031] Table 2 Optimal cutoff index table ; In the present embodiment, in order to test the diagnostic model of oligoasthenozoospermia constructed above, 70% of the entire metabolomics samples were used as a training set to train the model, and then the trained model was used to predict the remaining 30% of the metabolomics samples. ROC analysis was performed based on the expression levels of the two biomarkers in the samples. The results are as follows: Fig.11 As shown in the figure, the AUC value of the training set is 0.998, and the AUC value of the test set is 0.963. This shows that the model of candidate biomarker logistic regression has a good effect on the classification of samples in the test set.
[0032] This method uses non-targeted metabolomics to systematically analyze seminal plasma metabolites. As a result, a large number of metabolites and metabolic pathways related to oligoasthenozoospermia were identified, providing a comprehensive perspective on the changes in metabolites in patients with oligoasthenozoospermia. Using high-resolution data sets, a diagnostic model was constructed to distinguish oligoasthenozoospermia from normal sperm. At the same time, two biomarkers, 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol, were discovered and validated, which can be used to prepare kits and detection equipment for identifying male oligoasthenozoospermia.
[0033] The above contents are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for screening biomarkers of oligoasthenozoospermia, characterized in that: The steps include the following: Step 1: Comparison of seminal plasma metabolomics between oligoasthenozoospermia patients and normal control group: Seminal plasma samples from oligoasthenospermia patients and normal men were collected for non-targeted metabolomics testing. According to the significance criteria of multidimensional statistical analysis and univariate statistical analysis, i.e., VIP>1, p When value is less than 0.05, the significant difference of metabolites between the two groups was statistically analyzed to screen candidate biomarkers; Step 2: Biomarker screening for oligoasthenozoospermia: By integrating machine learning methods, the comprehensive weight value of each metabolite was calculated, and the candidate biomarkers were ranked according to their comprehensive weight values. The ROC was used to evaluate the influence of each metabolite on the AUC value of the model. The top two metabolites in the AUC cumulative curve weight of the candidate biomarker, namely 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol, were selected to form a candidate biomarker combination. Step 3: Verify and characterize candidate biomarkers: The selected marker combinations were verified and evaluated by three methods: logistic regression, random forest, and support vector machine; Step 4: Build a diagnostic model through logistic regression to distinguish oligoasthenozoospermia from normal sperm, and evaluate the diagnostic ability of the diagnostic model.
2. The method for screening oligoasthenozoospermia biomarkers according to claim 1, characterized in that: In step 4, the biomarker logistic regression model formula constructed is: ； in, z The expression is as follows: ； Substitute the expression level values of the two biomarkers in the sample to be classified into the above model formula to calculate the probability value P ,if P If the value exceeds the cutoff value, it is classified as the oligoasthenozoospermia group, otherwise it is classified as the normal group.
3. An application of a biomarker for oligoasthenozoospermia, based on a biomarker combination screened by the method for screening a biomarker for oligoasthenozoospermia according to any one of claims 1 and 2, characterized in that: The biomarker combination, i.e., 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine and [6]-gingerol, is used to prepare a kit and a detection device for identifying male oligoasthenozoospermia.