Bile acid markers for risk stratification of liver fibrosis in women with sarcopenia, predictive models and use thereof

Abstract

The application discloses a set of bile acid marker combinations, kits and applications for female sarcopenia patients with liver fibrosis risk stratification. The bile acid marker combinations are GCA, GDCA, GCDCA, TDCA, TDCA-3S, TCDCA, GLCA-3S and GDCA-3S in serum or plasma. Through quality control, difference screening and multi-model verification, a joint scoring model is constructed to distinguish the basic bile acid metabolism type and the liver fibrosis susceptible type subgroup. The experimental results show that the AUC of the model verification queue reaches 0.982; the median of the susceptible type subgroup LSM is 6.0 kPa, and the high-risk proportion is 20.7%, which is significantly higher than that of the basic type. The application can realize early and accurate stratification of the liver fibrosis risk of female sarcopenia patients, and is especially suitable for clinical screening and risk assessment.

Description

Technical Field

[0001] This invention relates to a combination of bile acid biomarkers for risk stratification of liver fibrosis in female sarcopenia patients, as well as corresponding detection kits, predictive models and their applications, belonging to the field of biomarker detection technology. Background Technology

[0002] Sarcopenia is a clinical syndrome characterized by decreased skeletal muscle mass, muscle strength, and physical decline. It is more common in middle-aged and elderly people, and female patients often experience accompanying problems such as aging, changes in body fat distribution, glucose and lipid metabolism disorders, and increased hepatic metabolic burden. For female sarcopenia patients, early identification and risk stratification of liver fibrosis are of great significance in delaying the progression of liver disease and guiding clinical intervention.

[0003] Liver stiffness measurement (LSM) is currently a key indicator for non-invasive assessment of the degree of liver fibrosis. Elevated LSM indicates an increased risk of liver fibrosis occurrence and progression. Existing clinical risk assessment systems for female sarcopenia patients mostly rely on grip strength, skeletal muscle mass, routine biochemical indicators, or liver elasticity tests themselves, lacking a combination of specific metabolic biomarkers that can be directly detected from serum or plasma and achieve early risk stratification.

[0004] Currently, clinical assessment of liver fibrosis risk in female sarcopenia patients mainly relies on liver stiffness measurement, routine biochemical indicators, and muscle mass / strength assessment. While LSM testing is the non-invasive gold standard, it relies on specialized equipment, is costly, and is difficult to implement for large-scale early screening. Routine biochemical indicators lack specificity and cannot effectively distinguish between low-risk and high-risk groups. Sarcopenia-related indicators only reflect muscle status and have a weak correlation with liver fibrosis risk, making them unsuitable for precise stratification of liver lesions.

[0005] Metabolomics studies have shown that abnormal bile acid profiles are closely related to the occurrence of liver fibrosis and are expected to become early non-invasive biomarkers. However, existing bile acid-related studies have significant shortcomings: First, there is a lack of bile acid biomarker screening systems specifically for female sarcopenia patients, and biomarkers screened in the general population have poor stability and low efficacy in this population; second, single bile acid indicators have low specificity and are easily affected by individual metabolic fluctuations, making it difficult to achieve accurate stratification; third, existing models have not undergone rigorous quality control, multi-model stability validation, and clinical outcome (LSM≥8.0kPa) validation, resulting in insufficient predictive accuracy and poor reproducibility; fourth, there are currently no reagent kits or standardized predictive models that can be directly used in clinical practice, making it impossible to achieve rapid and convenient risk stratification assessment.

[0006] Therefore, developing specific combinations of bile acid biomarkers and risk assessment tools suitable for this population has significant clinical application value and is a practical need. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the primary technical problem this invention aims to solve is to provide a combination of bile acid biomarkers for risk stratification of liver fibrosis in female patients with sarcopenia. This combination exhibits high specificity and stability, and can accurately distinguish between basal bile acid metabolizers and subgroups susceptible to liver fibrosis.

[0008] Another technical problem to be solved by the present invention is to provide a method for constructing a risk stratification prediction model for liver fibrosis using the above-mentioned bile acid biomarker combination, so as to realize the scientific screening of biomarkers and the standardized construction of the model, and ensure the stability, repeatability and prediction accuracy of the model.

[0009] Another technical problem to be solved by the present invention is to provide a detection kit for risk stratification of liver fibrosis in female sarcopenia patients.

[0010] Another technical problem to be solved by this invention is to provide the application of the above-mentioned bile acid biomarker combination and prediction model in liver fibrosis risk assessment and stratified management, so as to provide technical support for early warning and precise intervention of liver fibrosis in female sarcopenia patients.

[0011] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution: A bile acid biomarker combination for risk stratification of liver fibrosis in female sarcopenia patients, consisting of eight bile acid metabolites: glycocholic acid (GCA), glycodeoxycholic acid (GDCA), glycochenodeoxycholic acid (GCDCA), taurideoxycholic acid (TDCA), taurideoxycholic acid-3-sulfate (TDCA-3S), taurideoxycholic acid (TCDCA), glycocholic acid-3-sulfate (GLCA-3S), and glycodeoxycholic acid-3-sulfate (GDCA-3S).

[0012] A method for constructing a risk stratification prediction model for liver fibrosis in female sarcopenia patients includes the following steps: (1) Female patients diagnosed with sarcopenia were included. Serum or plasma samples were collected and the bile acid concentration in the samples was measured to obtain raw bile acid concentration data. (2) Remove bile acids with a missing rate > 30% and retain candidate bile acids that have passed quality control; (3) Based on the candidate bile acids, after standardization and clustering, the basal bile acid metabolizer and liver fibrosis susceptibility subgroups were divided; (4) The eight bile acid biomarkers described in claim 1 were determined through statistical difference screening, effect size screening, multi-model stability evaluation and clinical outcome association verification. (5) Calculate model parameters based on the screening queue data, construct a joint scoring model and determine the risk threshold to complete the model construction.

[0013] Preferably, in step (4), the statistical difference screening adopts the P<0.05 standard, and the multi-model stability evaluation adopts at least one algorithm among LASSO, Gradient Boosting Machine (GBM), Support Vector Machine (SVM), Random Forest, and Decision Tree.

[0014] Preferably, in step (5), the joint scoring model includes: Standardization: z i =(C i -μ i ) / σ i ; Joint scoring model: S = b + Σ(w) i ×z i ); Risk probability: p = 1 / (1 + exp(-S)); Where C i Let μ be the concentration of the i-th bile acid in the sample to be tested. i σ i w represents the mean / standard deviation of the training set. i is the model weight for the i-th bile acid biomarker, and b is the model intercept.

[0015] Preferably, in step (5), the risk threshold T is determined by the Youden index of the ROC curve of the screening queue; the judgment rule is: when p≥T, it is judged as a liver fibrosis susceptible type, and when p<T, it is judged as a basic bile acid metabolism type.

[0016] A risk stratification prediction model for liver fibrosis in female sarcopenia patients was constructed using the method described above.

[0017] A diagnostic kit for risk stratification of liver fibrosis in female patients with sarcopenia, comprising at least: (1) Standards, internal standards, and quality control products for detecting the above 8 bile acid metabolites; (2) Serum / plasma sample pretreatment reagents; (3) Record the instruction manual for the above model parameters, thresholds and interpretation criteria.

[0018] A data processing system for risk stratification of liver fibrosis in female sarcopenia patients, comprising at least: (1) Bile acid concentration data input module; (2) Data standardization and joint scoring calculation module; (3) Automatic interpretation module for liver fibrosis susceptibility subgroups; (4) LSM≥8.0 kPa High-risk association verification module.

[0019] The above-mentioned combination of bile acid biomarkers was used in the preparation of a risk stratification product for liver fibrosis in female sarcopenia patients. By detecting the concentration of eight bile acids, the product distinguishes between basal bile acid metabolizers and liver fibrosis susceptibility subgroups.

[0020] The above predictive model was used in the risk assessment of liver fibrosis in female sarcopenia patients, with LSM≥8.0 kPa as the high-risk standard for liver fibrosis, to predict the high-risk probability of patients and manage them in a stratified manner.

[0021] Compared with the prior art, the present invention has the following technical effects: 1. The bile acid biomarker combinations provided by this invention exhibit significant differences and strong discriminative ability. Screening cohort results from this invention show that the levels of all eight bile acid biomarkers in the liver fibrosis susceptibility subgroup were significantly higher than those in the basal bile acid metabolizer subgroup. Among them, the median fold changes for GDCA-3S, GDCA, and TDCA reached 5.14, 6.13, and 5.71, respectively, with P values ​​all less than 1 × 10⁻⁶. -6 The differences are extremely significant, which can effectively distinguish between the two metabolic subgroups.

[0022] 2. The prediction model provided by this invention exhibits excellent discriminative power and high stability. The joint scoring model constructed based on eight bile acid biomarkers achieved an AUC of 1.000 in the screening cohort and 0.982 in the validation cohort, with a sensitivity of 0.929, specificity of 1.000, and accuracy of 0.962. Validation with multiple models including LASSO, GBM, SVM, and random forest all showed AUCs higher than 0.96, demonstrating the model's strong discriminative ability and good cross-algorithm stability.

[0023] 3. This invention is highly correlated with the clinical outcomes of liver fibrosis and exhibits precise stratification. Using an LSM ≥ 8.0 kPa as the high-risk criterion for liver fibrosis, the median LSM in the susceptible subgroup of the overall cohort was 6.0 kPa, significantly higher than the baseline 4.4 kPa; the proportion of the susceptible subgroup with LSM ≥ 8.0 kPa reached 20.7%, far exceeding the baseline 5.3%; in the screening cohort, GCA, GDCA, GCDCA, and TDCA were significantly correlated with LSM ≥ 8.0 kPa, with P values ​​of 0.009, 0.013, 0.018, and 0.043, respectively, clearly reflecting the risk of liver fibrosis.

[0024] 4. The present invention features rigorous quality control, standardized screening procedures, and strong reproducibility. Of the original 58 bile acids in this invention, 28 candidates were retained after elimination of those with a deficiency rate >30%. These were then subjected to initial screening (P < 0.05), directional consistency screening, and multi-model stability verification, ultimately identifying 8 biomarkers. The screening process was transparent, with clear quality control standards, ensuring stable and reproducible results.

[0025] 5. This invention is non-invasive and convenient, making it particularly suitable for clinical screening scenarios. The invention only requires the collection of serum or plasma samples to complete the test. The accompanying kit provides standardized pretreatment reagents, standards, internal standards, and interpretation instructions. It is simple to operate, minimally invasive, and can be used for early screening, stratified management, and dynamic monitoring of liver fibrosis risk in female sarcopenia patients, demonstrating strong clinical applicability. Attached Figure Description

[0026] Figure 1 This is a flowchart of the overall technical solution of the present invention; Figure 2 Flowchart for bile acid quality control and biomarker screening; Figure 3 A graph showing the differences in eight bile acid biomarkers among basal bile acid metabolizers and liver fibrosis-susceptible bile acid metabolizer subgroups. Figure 4 ROC curves for identifying bile acid metabolic subgroups susceptible to liver fibrosis using a combined scoring model of eight bile acids; Figure 5A A comparison of the median LSM values ​​for different bile acid metabolism subsets; Figure 5B A comparison chart of the proportions of LSM≥8.0 kPa for different bile acid metabolism subgroups; Figure 6 The ROC curve is a supplementary prediction of the high-risk status of liver fibrosis with LSM≥8.0 kPa using a combined scoring model of eight bile acids. Detailed Implementation

[0027] The embodiments of this invention are for illustrative purposes only and are not intended to limit the scope of the invention. Experimental methods in the following embodiments, unless otherwise specified, are generally performed under conventional conditions or as recommended by the manufacturer. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of skill in the art. Furthermore, any methods and materials similar or equivalent to those described herein may be applied to this invention. The preferred embodiments and materials described herein are for illustrative purposes only.

[0028] Figure 1 The flowchart of the overall technical solution of the present invention shows the process from sample collection from female sarcopenia patients, bile acid detection, quality control of missing rate, screening cohort screening, determination of 8 biomarkers, establishment of joint scoring model, validation cohort validation to LSM outcome validation. Figure 2 The flowchart for bile acid quality control and biomarker screening shows that after removing 58 original bile acids with a deletion rate >30%, 28 candidate bile acids were retained. These were then further screened using a P < 0.05 ratio, directional consistency, and machine learning stability to obtain 8 biomarkers. Based on this, the bile acid biomarker combination screening and model construction methods are as follows.

[0029] Example 1: Screening of bile acid biomarker combinations 1. Sample Acquisition and Bile Acid Detection The female sarcopenic patients referred to in this invention are female individuals diagnosed with sarcopenia based on clinical body composition measurements and muscle strength evaluations. Sarcopenia can be diagnosed according to the Asian Sarcopenia Working Group criteria. A total of 84 female sarcopenic patients were included, and serum or plasma samples were obtained from the subjects. Targeted bile acid detection technology was used to measure the levels of bile acid metabolites. The detection subjects included the original 58 bile acid metabolites. LC-MS / MS technology was preferably used for bile acid detection, and quantification was performed using the internal standard method.

[0030] 2. Quality control of candidate bile acids Quality control screening was performed on 58 original bile acid metabolites. Thirty bile acid metabolites with a deletion rate greater than 30% in female sarcopenia samples were removed, leaving 28 bile acid metabolites with a deletion rate of no more than 30% as candidate bile acids. This deletion rate quality control step is crucial for excluding bile acid markers with unstable detection or insufficient reproducibility, and is an important step in marker screening.

[0031] 3. Construction of the filtering queue and validation queue This study included 84 female patients with sarcopenia. Based on LSM test results, the subjects were divided into three strata: LSM < 8.0 kPa, LSM ≥ 8.0 kPa, and no LSM test results. Subsequently, within each stratum, they were randomly assigned to a screening cohort and a validation cohort at a ratio of approximately 7:3.

[0032] In one embodiment of the present invention, a "screening queue" refers to a set of samples used for biomarker discovery, model training, and threshold determination; a "validation queue" refers to a set of samples that does not participate in biomarker screening and threshold determination, but is only used to validate the efficacy of a fixed biomarker combination. The validation queue is derived from a stratified division of the same research population and is an internal validation queue.

[0033] In this embodiment, the screening queue consisted of 58 cases, and the validation queue consisted of 26 cases. In the screening queue, 41 cases had LSM < 8.0 kPa, 5 cases had LSM ≥ 8.0 kPa, and 12 cases had no LSM data. In the validation queue, 18 cases had LSM < 8.0 kPa, 3 cases had LSM ≥ 8.0 kPa, and 5 cases had no LSM data.

[0034] 4. Definition of bile acid metabolism subsets Based on candidate bile acid metabolites, bile acid profiles were stratified to classify female sarcopenia patients into basal bile acid metabolizers and liver fibrosis-susceptible bile acid metabolizer subgroups. The basal bile acid metabolizer subgroup was characterized by an overall low bile acid profile and a low risk of liver fibrosis; the liver fibrosis-susceptible bile acid metabolizer subgroup was characterized by an overall elevated bile acid profile, particularly elevated conjugated, sulfated, and secondary-related bile acids, accompanied by a higher risk of LSM (Low-Symptom Complex). Specifically, in the screening cohort, candidate bile acids with a missing rate of no more than 30% underwent log1p transformation and Z-score standardization, and binary classification was performed based on the standardized bile acid matrix. After stratification, the levels of total bile acids, glycine-conjugated bile acids, taurine-conjugated bile acids, sulfated bile acids, and secondary-related bile acids, as well as the median LSM and the proportion of LSM ≥ 8.0 kPa, were calculated for both subgroups. A subgroup with elevated overall bile acid levels, especially glycine-conjugated bile acids, taurine-conjugated bile acids, sulfated bile acids, and secondary-related bile acids, accompanied by higher LSM levels or a higher proportion of LSM ≥ 8.0 kPa, is defined as the liver fibrosis-susceptible bile acid metabolism subgroup. Another subgroup is defined as the basic bile acid metabolism type, which is a metabolic phenotype with an overall lower bile acid profile and a lower risk of liver fibrosis.

[0035] In the screening cohort, there were 36 cases of basal bile acid metabolism and 22 cases of bile acid metabolism subgroup susceptible to liver fibrosis; in the validation cohort, there were 12 cases of basal bile acid metabolism and 14 cases of bile acid metabolism subgroup susceptible to liver fibrosis.

[0036] 5. Screening process for 8 bile acid biomarkers This embodiment uses a four-stage screening method to determine the final eight bile acid biomarkers. The screening process is shown in Table 1.

[0037] The first stage of screening was a quality control screening. The initial test included 58 bile acid metabolites; 30 metabolites with a deletion rate greater than 30% were removed, leaving 28 candidate bile acids.

[0038] The second level of screening was a preliminary screening based on statistical differences. In the screening cohort, participants were grouped into basal bile acid metabolism subgroups and those susceptible to liver fibrosis. Using P < 0.05 as the differential screening criterion, 18 differentially expressed bile acids were obtained from 28 candidate bile acids. The screening results are shown in Table 2.

[0039] The third level of screening involves directional consistency and effect size screening. Bile acid metabolites that show a directional increase in the bile acid metabolism subgroups susceptible to liver fibrosis and whose median fold change meets the preset criteria are retained. Conjugated bile acids, sulfated bile acids, and secondary related bile acids with biological explanatory value are given priority for retention.

[0040] The fourth level of screening involved machine learning stability evaluation and LSM outcome association evaluation. LASSO, GBM, SVM, decision trees, and random forests were used to evaluate the importance of candidate bile acids. Combined with the correlation between LSM ≥ 8.0 kPa and the high risk of liver fibrosis, bile acids that showed stable discriminative contributions in multiple models, were consistent with the statistical screening direction, and were associated with the high-risk LSM status were prioritized for retention.

[0041] like Figure 3 Table 3 shows the results of the final screening cohort differences for the eight bile acid biomarkers. The eight bile acid biomarkers were finally identified as: GCA, GDCA, GCDCA, TDCA, TDCA-3S, TCDCA, GLCA-3S, and GDCA-3S.

[0042] Table 1. Screening process for bile acid biomarkers Table 2. Candidate differentially expressed bile acids with P < 0.05 in the screening cohort. Table 3. Results of the screening cohort differences for the final eight bile acid biomarkers Example 2: Establishment of a Joint Scoring Model The parameters of the combined scoring model for 8 bile acids are shown in Table 4.

[0043] Detection of the concentrations of 8 bile acid markers C i , i = 1...8. The concentrations are standardized according to the screening queue parameters to obtain: z i =(C i -μ i ) / σi.

[0044] Where μ i To filter the mean of the i-th bile acid marker in the queue, σ i To select the standard deviation of the i-th bile acid marker in the screening cohort, the standardized bile acid values ​​are substituted into the joint scoring model: S = b + Σ(w i ×z i ).

[0045] Further calculate the risk probability: p = 1 / (1 + exp(-S)). Where w i Let be the model weight for the i-th bile acid biomarker, and b be the model intercept. The threshold T is determined based on the Youden index according to the ROC curve of the screening cohort. When p ≥ T, the subgroup is identified as susceptible to liver fibrosis and a bile acid metabolism subgroup; when p < T, the subgroup is identified as basal bile acid metabolism.

[0046] Table 4. Parameters of the combined scoring model for 8 bile acids Example 3: Verification Queue Verification The eight bile acid biomarkers identified in the screening cohort, μ i σ i w i After fixing b and the threshold T, the results were applied to the validation cohort. The validation cohort did not participate in biomarker screening, model weight determination, or threshold adjustment. The results were used to evaluate the stability of the combination of the eight bile acid biomarkers in identifying bile acid metabolism subgroups susceptible to liver fibrosis.

[0047] like Figure 4 As shown in Table 5, the AUC of the screening cohort for the combined scoring model of eight bile acids to identify bile acid metabolic subgroups susceptible to liver fibrosis was 1.000, and the AUC of the validation cohort was 0.982, indicating that the combination of these eight bile acid biomarkers still maintained a high discriminative ability in the validation cohort.

[0048] Table 5 Performance of the combined scoring model for 8 bile acids As shown in Table 6, five machine learning models were used to evaluate the discriminative stability of candidate bile acids for bile acid metabolic subpopulations. The results showed that LASSO, GBM, SVM, and Random Forest all had high subpopulation discrimination AUCs in the validation cohort, suggesting that this combination of bile acid biomarkers is not dependent on a single algorithm but has cross-model stability.

[0049] Table 6. Validation results of the stability of five machine learning models in discriminating bile acid metabolic subsets. Example 4: LSM Outcome Verification This invention uses LSM ≥ 8.0 kPa as the cutoff value for a high-risk state of liver fibrosis. LSM is the liver stiffness value obtained from liver elasticity testing, which can be used for non-invasive assessment of liver fibrosis risk. Considering that LSM thresholds may vary among different etiologies, testing devices, and populations, LSM ≥ 8.0 kPa in this invention is not used as a pathological diagnostic criterion for liver fibrosis, but rather as a risk stratification threshold for a high-risk state of liver fibrosis or a state requiring further evaluation in female sarcopenia patients. The selection of 8.0 kPa as the threshold is based on a commonly used interval in clinical elasticity testing for early screening and stratified management of liver fibrosis risk. Using LSM ≥ 8.0 kPa as a high-risk state of liver fibrosis, the LSM levels and the proportion of LSM ≥ 8.0 kPa are compared between the baseline bile acid metabolism subgroup and the bile acid metabolism subgroup susceptible to liver fibrosis.

[0050] Table 7. Validation of LSM outcomes for different bile acid metabolism subsets like Figure 5A , Figure 5B As shown in Table 7, the bile acid metabolism subgroup susceptible to liver fibrosis had a higher median LSM and a higher proportion of LSM ≥ 8.0 kPa in the overall cohort, screening cohort, and validation cohort, suggesting that this subgroup has a higher risk of liver fibrosis.

[0051] In individuals with LSM data, GCA, GDCA, GCDCA, and TDCA were significantly associated with a high risk of liver fibrosis at LSM ≥ 8.0 kPa, further supporting the association between the final combination of eight bile acid biomarkers and the risk of liver fibrosis.

[0052] Table 8. Bile acids associated with high risk of liver fibrosis with LSM ≥ 8.0 kPa Example 5: A supplementary model for direct LSM prediction In one embodiment of the present invention, the combined score of eight bile acids and / or the label of the bile acid metabolic subgroup susceptible to liver fibrosis are further input into the LSM risk assessment module, which outputs the risk probability of LSM ≥ 8.0 kPa. This module, as a preferred embodiment, is used to further illustrate the predictive relationship between the combination of eight bile acid biomarkers and the high-risk state of LSM.

[0053] like Figure 6 As shown, when the combined score of eight bile acids directly predicted LSM ≥ 8.0 kPa, the AUC of the screening cohort was 0.834, and the AUC of the validation cohort was 0.667. Since the number of positive cases with LSM ≥ 8.0 kPa in the validation cohort was small, this result is considered a supplementary example and does not constitute the sole core conclusion of this invention. The core technical solution of this invention is to identify a subgroup of bile acid metabolism susceptible to liver fibrosis using eight bile acid biomarkers, and to verify that this subgroup has a higher risk of liver fibrosis through LSM outcomes.

[0054] This invention establishes a screening and validation process for bile acid biomarkers based on 84 female patients with sarcopenia. First, 30 bile acids with a loss rate greater than 30% were removed from 58 original bile acid metabolites, retaining 28 candidate bile acids. Subsequently, a preliminary screening was conducted in the screening cohort using a p-value < 0.05, resulting in 18 differentially expressed bile acids. Further evaluation using multiple models including LASSO, GBM, SVM, decision trees, and random forests, combined with directional consistency, fold change, outcome association with LSM ≥ 8.0 kPa, and stability assessment, ultimately identified eight bile acid biomarkers: GCA, GDCA, GCDCA, TDCA, TDCA-3S, TCDCA, GLCA-3S, and GDCA-3S.

[0055] An eight-bile acid combined scoring model can stably identify a subgroup of bile acid metabolism susceptible to liver fibrosis in female sarcopenia patients. Validation cohort results show that the fixed combination of eight bile acid biomarkers maintains high discriminative power without rescreening or threshold readjustment, indicating that this biomarker combination has good reproducibility and translational potential.

[0056] Using LSM ≥ 8.0 kPa as a high-risk outcome for liver fibrosis, the fibrosis-susceptible bile acid metabolism subgroup showed higher LSM levels and a higher proportion of LSM ≥ 8.0 kPa compared to the baseline bile acid metabolism subgroup. These results indicate that this combination of eight bile acid biomarkers can not only identify subgroups with abnormal bile acid metabolism but also be used for risk stratification of liver fibrosis in female sarcopenia patients.

Claims

1. A combination of bile acid biomarkers for risk stratification of liver fibrosis in female patients with sarcopenia, characterized in that... It consists of eight bile acid metabolites: glycocholic acid (GCA), glycodeoxycholic acid (GDCA), glycochenodeoxycholic acid (GCDCA), taurideoxycholic acid (TDCA), taurideoxycholic acid-3-sulfate (TDCA-3S), taurideoxycholic acid (TCDCA), glycocholic acid-3-sulfate (GLCA-3S), and glycodeoxycholic acid-3-sulfate (GDCA-3S).

2. A method for constructing a risk stratification prediction model for liver fibrosis in female sarcopenia patients, characterized in that... include: (1) Female patients diagnosed with sarcopenia were included. Serum or plasma samples were collected and the bile acid concentration in the samples was measured to obtain raw bile acid concentration data. (2) Remove bile acids with a missing rate > 30% and retain candidate bile acids that have passed quality control; (3) Based on the candidate bile acids, after standardization and clustering, the basal bile acid metabolizer and liver fibrosis susceptibility subgroups were divided; (4) The eight bile acid biomarkers described in claim 1 were determined through statistical difference screening, effect size screening, multi-model stability evaluation and clinical outcome association verification. (5) Calculate model parameters based on the screening queue data, construct a joint scoring model and determine the risk threshold to complete the model construction.

3. The construction method as described in claim 2, characterized in that... In step (4), the statistical difference screening adopts the P<0.05 standard, and the multi-model stability evaluation adopts at least one algorithm among LASSO, Gradient Boosting Machine (GBM), Support Vector Machine (SVM), Random Forest, and Decision Tree.

4. The construction method as described in claim 2, characterized in that... In step (5), the joint scoring model includes: Standardization: z i =(C i -μ i ) / σ i ; Joint scoring model: S = b + Σ(w) i ×z i ); Risk probability: p = 1 / (1 + exp(-S)); Among them, C i Let μ be the concentration of the i-th bile acid in the sample to be tested. i σ i w represents the mean / standard deviation of the training set. i is the model weight for the i-th bile acid biomarker, and b is the model intercept.

5. The construction method as described in claim 2, characterized in that... In step (5), the risk threshold T is determined by the Youden index of the ROC curve of the screening cohort; the judgment rule is: when p≥T, it is judged as a liver fibrosis susceptible type, and when p<T, it is judged as a basic bile acid metabolism type.

6. A risk stratification prediction model for liver fibrosis in female sarcopenic patients, characterized in that... It is constructed by the construction method described in any one of claims 2 to 5.

7. A diagnostic kit for risk stratification of liver fibrosis in female patients with sarcopenia, characterized in that... Include: (1) Standards, internal standards, and quality control products for detecting the eight bile acid metabolites described in claim 1; (2) Serum / plasma sample pretreatment reagents; (3) A description of the model parameters, thresholds and interpretation criteria described in claim 6.

8. A data processing system for risk stratification of liver fibrosis in female sarcopenia patients, characterized in that... include: (1) Bile acid concentration data input module; (2) Data standardization and joint scoring calculation module; (3) Automatic interpretation module for liver fibrosis susceptibility subgroups; (4) LSM≥8.0 kPa High-risk association verification module.

9. Use of the bile acid marker combination of claim 1 in the preparation of a risk stratification product for liver fibrosis in female sarcopenic patients.

10. Use of the predictive model of claim 6 in assessing the risk of liver fibrosis in female patients with sarcopenia.

Images