Application of urine polypeptide in preparation of early detection product for renal function progress, model and method

Abstract

The invention relates to the field of medical treatment, in particular to application, a model and a method of urine polypeptide in preparation of a product for early detection of renal function progression, and the mass-to-charge ratio of the urine polypeptide is any one or more of m / z 1378.8, m / z 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z 2752.42, m / z 1013.59, m / z 1235.53, m / z 1911.76, m / z 2039.86, m / z 1949.9, m / z 1869.59, m / z 2086.71, m / z 1641.2 and m / z 3321.3. A model constructed according to the urine polypeptide can be used as a non-invasive tool for early detection and monitoring of kidney diseases, and renal insufficiency can be identified and indicated before a traditional clinical biomarker is abnormal by analyzing a unique mode of the polypeptide existing in a urine sample, so that intervention can be performed earlier.

Description

Technical Field

[0001] The present invention relates to the medical field, and in particular to the use, model and method of urine polypeptides in preparing early detection products for renal function progression. Background Art

[0002] Abnormal renal function is a major public health issue that affects millions of people worldwide and represents a significant burden on healthcare systems. Despite its high prevalence, chronic kidney disease (CKD) is often diagnosed only at an advanced stage due to the lack of specific symptoms in the early stages of the disease. Early detection and detection or prediction of the progression of abnormal renal function is essential for timely intervention in community populations, which can help prevent further deterioration of renal function and reduce the risk of complications such as end-stage renal disease (ESRD).

[0003] Despite numerous attempts to develop prediction models for the progression of chronic kidney disease (CKD) worldwide, most of these models have focused on European and North American populations. In 2011, Levey and colleagues reported a model developed and validated in a Canadian population that accurately predicted the progression of CKD from stage 3 to stage 5 to renal failure using routine laboratory results. Subsequently, this formula was extended to 31 cohorts covering four continents and more than 30 countries. The results showed that although the model showed high discrimination and adequate calibration in the North American cohort, its performance in non-North American regions needed to be optimized by adding recalibration factors. These findings suggest that racial differences limit the generalizability of prediction formulas in different populations. As described in these studies, existing studies have focused primarily on the progression from eGFR G3 to G5 to end-stage renal failure (requiring dialysis or transplantation). However, research on early recognition and progression of renal impairment remains insufficient.

[0004] Urine peptide analysis is a promising non-invasive tool for early detection and monitoring of kidney disease. By analyzing the unique patterns of peptides present in urine samples, it is possible to identify biomarkers indicative of renal dysfunction before traditional clinical biomarkers such as estimated glomerular filtration rate (eGFR) or proteinuria become abnormal. Previous studies have demonstrated the effectiveness of urine peptide analysis in distinguishing patients with chronic kidney disease from healthy people and predicting disease progression in clinical settings.

[0005] However, the applicability and effectiveness of urine peptide analysis for early detection and prediction of kidney disease progression in community populations, especially when conventional CKD screening methods are limited, still need more research to verify. This means that although this method shows some promise in specific clinical settings, further research is needed to determine its reliability and practical application value when used in a wider community population. Summary of the invention

[0006] In view of the shortcomings of the prior art described above, the purpose of the present invention is to provide the present invention relates to the medical field, in particular to the use, model and method of urine polypeptides in the preparation of early detection products for renal function progression, so as to solve the problems in the prior art.

[0007] To achieve the above-mentioned purpose and other related purposes, the present invention provides the use of urine polypeptides in preparing early detection products for progression of renal function, wherein the mass-to-charge ratio of the urine polypeptide is any one or more of the following, and the mass-to-charge ratio of the urine polypeptide is as follows, and the allowable deviation range of the mass-to-charge ratio is 500ppm: m / z 1378.8, m / z 1725.95, m / z1902.7, m / z 1933.8, m / z 2234.26, m / z 2752.42, m / z 1013.59, m / z 1235.53, m / z1911.76, m / z 2039.86, m / z 1949.9, m / z 1869.59, m / z 2086.71, m / z 1641.2, and m / z 3321.3.

[0008] The present invention also provides a product for early detection of renal function progression, wherein the product comprises any one or more of the urine polypeptides, or comprises an antibody that can specifically bind to the urine polypeptides.

[0009] The present invention also provides a method for constructing an early detection model for renal function progression, the method comprising the following steps:

[0010] S1, obtaining relative peak intensity data of urine peptides with the following mass-to-charge ratios of normal renal function population and abnormal renal function population: m / z 1378.8, m / z 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z2752.42, m / z 1013.59, m / z 1235.53, m / z 1911.76, m / z 2039.86, m / z 1949.9, m / z1869.59, m / z 2086.71, m / z 1641.2, m / z 3321.3; and using the obtained relative peak intensity data of each urine peptide as a training data set to construct a machine learning model, and directly or indirectly obtain the urine peptide index of the sample to be tested according to the constructed machine learning model;

[0011] S2, using the urine polypeptide index obtained in S1 as the independent variable, and whether the primary or secondary endpoint event occurs as the dependent variable to construct a logistic regression model, the logistic regression model is used for the early detection of renal function progression, the primary endpoint event refers to a significant decrease in the renal disease index of the sample to be tested when the predetermined time is reached; the secondary endpoint event refers to the sample to be tested reaching the renal disease diagnostic criteria when the predetermined time is reached.

[0012] The present invention also provides an early detection model for renal function progression obtained by the model construction method.

[0013] The present invention also provides a method for early detection of renal function progression, the method comprising the following steps:

[0014] 1) Calculation of urine polypeptide index: Input the relative peak intensity data of urine polypeptides with the following mass-to-charge ratios of the sample to be tested into the machine learning model to directly or indirectly obtain the urine polypeptide index of the sample to be tested: m / z 1378.8, m / z

[0015] 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z 2752.42, m / z 1013.59, m / z1235.53, m / z 1911.76, m / z 2039.86, m / z 1949.9, m / z 1869.59, m / z 2086.71, m / z1641.2, m / z 3321.3;

[0017] 2) Early detection of renal function progression and output of results: the urine polypeptide index obtained in step 1) is input into the early detection model of renal function progression, and the model outputs the result of whether a primary endpoint event or a secondary endpoint event occurs, wherein the primary endpoint event refers to a significant decrease in the renal disease index of the sample to be tested when a predetermined time is reached;

[0018] The secondary endpoint event refers to the test sample reaching the diagnostic criteria for kidney disease when the predetermined time is reached.

[0019] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned early detection method for renal function progression.

[0020] The present invention also provides a computer processing device, comprising a processor and the aforementioned computer-readable storage medium, wherein the processor executes a computer program on the computer-readable storage medium to implement the steps of the aforementioned method for early detection of renal function progression.

[0021] The present invention also provides an electronic terminal, comprising: a processor, a memory and a communicator; the memory is used to store computer programs, the communicator is used to communicate with external devices, and the processor is used to execute the computer program stored in the memory, so that the terminal performs an early detection method for renal function progression.

[0022] The present invention also provides a computer program product, comprising a computer program, which implements the steps of the method for early detection of renal function progression when executed by a processor.

[0023] As described above, the present invention relates to the medical field, and in particular to the use, model and method of urine polypeptides in the preparation of early detection products for the progression of renal function, and has the following beneficial effects: urine polypeptides are multi-parameter detections that can simultaneously detect multiple biomarkers, providing a more comprehensive assessment of the functional status of various parts such as renal tubules and glomeruli. It can be used as a non-invasive tool for early detection and monitoring of kidney disease. By analyzing the unique pattern of polypeptides present in urine samples, it is possible to identify indicators of renal insufficiency before abnormalities occur in traditional clinical biomarkers (such as estimated glomerular filtration rate eGFR or proteinuria), thereby intervening earlier. At the same time, the unique pattern of urine polypeptides identified may provide clues to the pathogenesis of kidney disease, help understand the pathological process of the disease and develop new therapeutic strategies for upstream pathway intervention of urine polypeptides. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 Shown are the intergroup changes in urine peptide peaks in the external independent validation population and the discovery cohort, where: Control, normal renal function group; eGFR G2, renal function between normal and abnormal, defined as 60mL / min / 1.73m 2 ≤ eGFR<90 mL / min / 1.73 m 2 and ACR < 30 mg / g; KD, abnormal renal function group, defined as eGFR < 60 mL / min / 1.73 m 2 Or ACR ≥ 30 mg / g; Method: bray, indicating that the PCOA dimensionality reduction method is bray-cutis algorithm; Derivation set 1, discovery cohort baseline data set; Validation set 1, external validation cohort 1; Validation set 2, external validation cohort 2; Figures F and G are distribution diagrams of the cohort population.

[0025] Figure 2The AUC results of various machine learning classification models established based on the 15 screened urine peptide peaks for different cohorts are shown, including: 10-cross validation, 10-fold cross validation of the discovery cohort training data set; validationset1, external validation cohort 1; Validation set2, external validation cohort 2. RF, random forest; SVM, support vector machine; NB, naive Bayes; KNN, K-nearest neighbor; GBM, gradient boosted decision tree; ctree, decision tree; adaboost, adaptive boosting. Figures A, B, and C are the results of the discovery cohort training population (480 control participants and 203 KD participants), external independent validation population 1 (96 control participants and 100 KD participants), and external independent validation population 2 (48 control participants and 194 biopsy-confirmed CKD patients), respectively.

[0026] Figure 3 Shown are the results of the association between baseline urine peptide index and primary endpoint events constructed using a random forest RF model.

[0027] Figure 4 Shown are the results of the association between baseline urine peptide index and secondary endpoint events constructed using a random forest RF model.

[0028] Figure 5 Shown is a schematic diagram of an electronic terminal. DETAILED DESCRIPTION

[0029] The present invention first provides the use of urine polypeptides in preparing products for predicting the progression of renal function, wherein the mass-to-charge ratio of the urine polypeptides is any one or more of the following: m / z 1378.8, m / z 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z 2752.42, m / z 1013.59, m / z 1235.53, m / z 1911.76, m / z 2039.86, m / z1949.9, m / z 1869.59, m / z 2086.71, m / z 1641.2, and m / z 3321.3, and the allowable deviation range of the mass-to-charge ratio is 500 ppm.

[0030] The "tolerance range of the mass-to-charge ratio is 500ppm" means that the fluctuation range of the mass-to-charge ratio is within ±0.5%. Taking m / z 1378.8 as an example, the mass tolerance range of m / z 1378.8 is m / z 1378.8±6.894Da. This range is the tolerance range on the time-of-flight mass spectrometer, so when the aforementioned mass-to-charge ratio is mentioned in the present invention, this tolerance range is included by default.

[0031] The prediction may be called early detection, which usually refers to detection in the early stages of chronic kidney disease (CKD). In this stage, kidney function begins to be impaired, but there are usually no obvious symptoms. Early detection and intervention are crucial to prevent the disease from progressing to a more serious stage.

[0032] According to the guidelines of the Kidney Disease International Organization (KDIGO), chronic kidney disease is divided into five stages (G1-G5) based mainly on the estimated glomerular filtration rate (eGFR) and the degree of proteinuria:

[0033] G1: normal or near-normal renal function (eGFR ≥ 90 mL / min / 1.73 m 2 ), but there is evidence of renal damage, such as proteinuria, hematuria, or abnormal imaging.

[0034] G2 stage: mild renal function decline (eGFR 60-89mL / min / 1.73m 2 ), with evidence of renal damage.

[0035] G3 stage: moderate renal function decline, divided into two sub-stages:

[0036] Stage G3a: eGFR 45-59mL / min / 1.73m 2

[0037] G3b: eGFR 30-44 mL / min / 1.73 m 2

[0038] G4 stage: severe renal function decline (eGFR 15-29mL / min / 1.73m 2 ).

[0039] G5: End-stage renal disease (eGFR < 15 mL / min / 1.73 m 2 ), requiring dialysis or a kidney transplant.

[0040] Early stages of kidney disease usually refer to G1 and G2 stages, which are characterized by the following:

[0041] G1 stage (normal or near normal renal function):

[0042] Evidence of renal injury: present, but renal function is essentially normal. Evidence of injury may include: proteinuria (increased urinary protein excretion), hematuria (increased red blood cells in the urine), imaging abnormalities (such as structural abnormalities shown on renal ultrasound, CT or MRI), and histopathological abnormalities (abnormalities found through renal biopsy).

[0043] G2 (mildly decreased renal function): Evidence of renal damage is similar to G1, with evidence of renal damage but mildly decreased renal function.

[0044] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1378.8 is shown in SEQ ID NO.1.

[0045] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1725.95 is shown in SEQ ID NO.2.

[0046] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1902.7 is shown in SEQ ID NO.3.

[0047] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1933.8 is shown in SEQ ID NO.4.

[0048] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2234.26 is shown in SEQ ID NO.5.

[0049] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2752.42 is shown in SEQ ID NO.6.

[0050] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1013.59 is shown in SEQ ID NO.7.

[0051] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1235.53 is shown in SEQ ID NO.8.

[0052] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1911.76 is shown in SEQ ID NO.9.

[0053] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2039.86 is shown in SEQ ID NO.10.

[0054] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1949.9 is shown in SEQ ID NO.11.

[0055] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1869.59 is shown in SEQ ID NO.12.

[0056] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2086.71 is shown in SEQ ID NO.13.

[0057] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1641.2 is shown in SEQ ID NO.14.

[0058] The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 3321.3 is shown in SEQ ID NO.15.

[0059] LFM*GKVVNPTQK (SEQ ID NO.1)

[0060] SEGLKLVDKFLEDVK (SEQ ID NO. 2)

[0061] KTQTAGVEERTKSQKPL(SEQ ID NO.3)

[0062] KERDEDAHYPVGVNAMT(SEQ ID NO.4)

[0063] TGNRNPGSSGTGGTATWKPGSSGP(SEQ ID NO.5)

[0064] VFLMIEQNTKSPLFMGKVVNPTQK(SEQ ID NO.6)

[0065] PGTEPLGSPGT (SEQ ID NO. 7)

[0066] QHTCSALAPPGP (SEQ ID NO.8)

[0067] SGSVIDQSRVLNLGPITR(SEQ ID NO.9)

[0068] SGSVIDQSRVLNLGPITRK(SEQ ID NO.10)

[0069] SGSVIDQSRVLNLGPITRK(SEQ ID NO.11)

[0070] YQKVVAGVANALAHKYH(SEQ ID NO.12)

[0071] DEAGSEADHEGTHSTKRGHA(SEQ ID NO.13)

[0072] ERVM*TRSYGATATSP(SEQ ID NO.14)

[0073] DLDFAKAKQLTVVGCQFTEFLLESEEDGQ(SEQ ID NO.15)

[0074] The product for predicting the progression of renal function is a product for predicting the progression of renal disease. In certain embodiments of the present invention, the renal disease is chronic kidney disease (CKD).

[0075] The present invention also provides a product for predicting the progression of renal function, wherein the product comprises any one or more of the urine polypeptides.

[0076] The present invention also provides a method for constructing a model for predicting renal function progression, the method comprising the following steps:

[0077] S1, obtaining relative peak intensity data of urine peptides with the following mass-to-charge ratios of normal renal function population and abnormal renal function population: m / z 1378.8, m / z 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z2752.42, m / z 1013.59, m / z 1235.53, m / z 1911.76, m / z 2039.86, m / z 1949.9, m / z1869.59, m / z 2086.71, m / z 1641.2, m / z 3321.3; and using the obtained relative peak intensity data of each urine peptide as a training data set to construct a machine learning model, and directly or indirectly obtain the urine peptide index of the sample to be tested according to the constructed machine learning model;

[0078] S2, using the urine polypeptide index obtained in S1 as the independent variable, and whether the primary or secondary endpoint event occurs as the dependent variable to construct a logistic regression model, the logistic regression model is used to predict the progression of renal function, the primary endpoint event refers to a significant decrease in the renal disease index of the sample to be tested when the predetermined time is reached; the secondary endpoint event refers to the sample to be tested reaching the renal disease diagnostic criteria when the predetermined time is reached.

[0079] In certain embodiments of the present invention, the normal renal function population and the abnormal renal function population in S1 are divided according to the estimated glomerular filtration rate (eGFR) and the urine protein to creatinine ratio (ACR), and the normal renal function population has an eGFR ≥ 90 mL / min / 1.73 m 2 ACR < 30 mg / g, eGFR < 60 mL / min / 1.73 m 2 Or ACR ≥ 30 mg / g. The estimated glomerular filtration rate (eGFR) can be obtained by the MDRD (Modification of Diet in Renal Disease) formula or the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) formula.

[0080] In certain embodiments of the present invention, the relative peak intensity data of urine polypeptides of the normal renal function group and the abnormal renal function group described in S1 are obtained by MALDI-TOF MS (matrix-assisted laser desorption time-of-flight mass spectrometry) detection.

[0081] In certain embodiments of the present invention, the machine learning model described in S1 is selected from any one or more of the following: support vector machine (SVM), naive Bayes (NB), random forest (RF), K-nearest neighbor (KNN), gradient boosted decision tree (GBM), decision tree (CTREE), adaptive boost (ADABOOST) or logistic regression. These machine learning models can be implemented by the Caret package (version 6.0-88) in R language.

[0082] In certain embodiments of the present invention, the machine learning model in S1 is a random forest model, and the random forest model is constructed in the following manner: using the relative peak intensity data of each urine polypeptide obtained as a training data set to perform model training, and performing an accuracy test, and obtaining the optimal model according to the accuracy test results as the urine polypeptide index calculation model.

[0083] When the random forest model is used, the relative peak intensity data of the urine polypeptides of the sample to be tested is input into the urine polypeptide index calculation model, and the model can directly output the urine polypeptide index of the sample to be tested.

[0084] The machine learning models of support vector machine (SVM), naive Bayes (NB), K-nearest neighbor (KNN), gradient boosted decision tree (GBM), decision tree (CTREE), and adaptive boost (ADABOOST) are constructed in the same way as the random forest, and can also directly output the urine polypeptide index of the sample to be tested.

[0085] In certain embodiments of the invention, the method of testing the accuracy of the model includes independent validation or cross validation.

[0086] In certain embodiments of the present invention, independent validation is to use an independent validation cohort other than the training data set (ie, the discovery cohort) to test the model effect.

[0087] In certain embodiments of the present invention, the cross validation is selected from k-fold cross validation. The k-fold cross validation can be three-fold, usually five or ten-fold. In a preferred embodiment, it is selected from ten-fold cross validation.

[0088] The accuracy of the model was evaluated using metrics such as AUC, accuracy, specificity, sensitivity, positive predictive value (PPV), negative predictive value (NPV), or F1 score.

[0089] In other embodiments of the present invention, the machine learning model in S1 is a logistic regression model, preferably, a univariate logistic regression model.

[0090] When the machine learning model in S1 is a logistic regression model, the urine polypeptide index can only be obtained indirectly according to the logistic regression model. Specifically, S1 includes the following steps:

[0091] S11, model construction: using the relative peak intensity data of each urine polypeptide obtained as the independent variable and whether there is renal function abnormality as the dependent variable to construct a logistic regression model;

[0092] S12, calculating the urine polypeptide index, according to the logistic regression model constructed in S11, obtaining the odds ratio (OR) of each urine polypeptide, substituting the OR value into formula ① to calculate the urine polypeptide index of the sample to be tested;

[0093]

[0094] Among them, UPI i represents the urine polypeptide index of the sample i to be tested,

[0095] n is the number of urine peptides, n = 15,

[0096] Intensity ij represents the relative peak intensity of the jth urine peptide in the sample i to be tested,

[0097] OR j represents the odds ratio of the jth urine peptide obtained by the logistic regression model.

[0098] In S12, the odds ratio OR is calculated by a logistic regression model, which indicates the multiple of the probability (odds) of the occurrence of the outcome event (or endpoint event) relative to the probability of not occurring when the independent variable increases by one unit (or specific change). The odds ratio is used to evaluate the strength of the association between a factor (such as an exposure factor) and an outcome (such as a disease). If the OR is greater than 1, it indicates that the independent variable is related to an increase in the outcome event; if the OR is less than 1, it indicates that the independent variable is related to a decrease in the outcome event; if the OR is equal to 1, it indicates that there is no association between the independent variable and the outcome event. According to the univariate logistic regression model constructed in step S11, it can be known whether a certain urine polypeptide is a protective factor or a risk factor for renal function. The protective factor of renal function refers to the relative peak intensity of a certain urine polypeptide fragment in the healthy group compared with the kidney disease group. It is higher than the abnormal renal function group. In this case, the urine polypeptide fragment is a protective factor, otherwise it is a risk factor.

[0099] In certain embodiments of the present invention, the logistic regression model in S2 is a univariate or multivariate logistic regression model.

[0100] The independent variables in the multivariate logistic regression model also include any one or more of the following confounding variables: age, gender, BMI (body mass index), hypertension, diabetes, estimated glomerular filtration rate (eGFR) or urine protein to creatinine ratio (ACR).

[0101] In certain embodiments of the present invention, the renal disease indicator in S2 is selected from eGFR and / or ACR.

[0102] In certain embodiments of the present invention, the significant decrease in the renal disease index in S2 is, for example, a decrease of the eGFR of the sample to be tested by more than 15%, more than 20% or more than 25% compared with the current value, and the value can be determined according to the actual situation. The "current" refers to the sampling time point of the data obtained in S1.

[0103] In certain embodiments of the present invention, the predetermined time in S2 refers to the number of years after the sampling time of the data acquired in S1, such as 1 year, 2 years, 3 years, 4 years, 5 years, 6 years or more than 7 years. Preferably, the predetermined time is 6 years.

[0104] The present invention also provides a model for predicting renal function progression obtained by the model construction method.

[0105] In certain embodiments of the present invention, the model for predicting renal function progression comprises:

[0106] Urine polypeptide index calculation module: including the machine learning model constructed in step S1 of the construction method; used to obtain relative peak intensity data of urine polypeptides of the mass-to-charge ratio of the sample to be tested as follows, and calculate the urine polypeptide index of the sample to be tested: m / z1378.8, m / z 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z 2752.42, m / z1013.59, m / z1235.53, m / z 1911.76, m / z 2039.86, m / z 1949.9, m / z 1869.59, m / z2086.71, m / z 1641.2, m / z3321.3;

[0107] Prediction module: includes the logistic regression model constructed in step S2 of the construction method, which is used to obtain the urine polypeptide index obtained by the urine polypeptide index calculation module and output the prediction result.

[0108] Since the model in this embodiment is basically the same as the principle of the aforementioned method embodiment, in the above method and model embodiments, the definitions, calculation methods, enumeration of implementation methods and enumeration of preferred implementation methods of the same features can be used interchangeably and will not be repeated here.

[0109] It should be noted that it should be understood that the division of the various modules of the above device is only a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. These modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, each module can be a separately established processing element, or it can be integrated in a certain chip for implementation. In addition, it can also be stored in a memory in the form of program code, and called and executed by a certain processing element. The function of the above database acquisition module. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in the processor element or instructions in the form of software.

[0110] For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more digital singnal processors (DSP), or one or more field programmable gate arrays (FPGA) or graphics processing units (GPU). For another example, when a module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processors that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0111] The present invention also provides a method for predicting renal function progression, the method comprising the following steps:

[0112] 1) Calculation of urine polypeptide index: Input the relative peak intensity data of urine polypeptides with the following mass-to-charge ratios of the sample to be tested into the machine learning model to directly or indirectly obtain the urine polypeptide index of the sample to be tested: m / z 1378.8, m / z

[0113] 1725.95, m / z 1902.7, m / z 1933.8, m / z 2234.26, m / z 2752.42, m / z 1013.59, m / z1235.53, m / z 1911.76, m / z 2039.86, m / z 1949.9, m / z 1869.59, m / z 2086.71, m / z1641.2, m / z 3321.3;

[0115] 2) predicting the progression of renal function and outputting the result: inputting the urine polypeptide index obtained in step 1) into the model for predicting the progression of renal function, and the model for predicting the progression of renal function outputs the result of whether a primary endpoint event or a secondary endpoint event occurs, wherein the primary endpoint event refers to a significant decrease in the renal disease index of the sample to be tested when a predetermined time is reached;

[0116] The secondary endpoint event refers to the test sample reaching the diagnostic criteria for kidney disease when the predetermined time is reached.

[0117] In certain embodiments of the present invention, if neither the primary endpoint event nor the secondary focus event occurs in step 2), it is considered that the renal function has not undergone progressive changes within the predetermined time.

[0118] In certain embodiments of the present invention, the significant decrease in the renal disease index in step 2) means that the eGFR of the sample to be tested decreases by more than 15%, more than 20% or more than 25% compared with the current value, and this value can be determined according to actual conditions.

[0119] The results of the model output in step 2) may also include: model description (such as general information such as the name, age, gender, whether the sample suffers from hypertension or diabetes, or an explanation of the meaning of the prediction results, etc.), preliminary suggestions on whether further diagnosis and treatment is needed, etc.

[0120] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the aforementioned method for predicting renal function progression when the program is executed by a processor.

[0121] The present invention also provides a computer processing device, comprising a processor and the aforementioned computer-readable storage medium, wherein the processor executes a computer program on the computer-readable storage medium to implement the steps of the aforementioned method for predicting renal function progression.

[0122] The present invention also provides an electronic terminal, comprising: a processor, a memory and a communicator; the memory is used to store computer programs, the communicator is used to communicate with external devices, and the processor is used to execute the computer program stored in the memory so that the terminal executes a method for predicting renal function progression.

[0123] like Figure 5 As shown, it is an optional hardware structure diagram of the electronic terminal 100 provided in an embodiment of the present invention. The terminal 100 can be a computer device, a tablet device, a personal digital processing device, a factory background processing device, etc. The electronic terminal 100 includes: at least one processor 101, a memory 102, at least one network interface 104 and a user interface 106. The various components in the device are coupled together through a bus system 105. It can be understood that the bus system 105 is used to realize the connection and communication between these components. In addition to the data bus, the bus system 105 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, in Figure 5 In the specification, various buses are labeled as bus systems.

[0124] The user interface 106 may include a display, a keyboard, a mouse, a trackball, a click gun, keys, buttons, a touch pad or a touch screen.

[0125] It is understood that the memory 102 can be a volatile memory or a non-volatile memory, and can also include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM). The memory described in the embodiments of the present invention is intended to include but is not limited to these and any other suitable categories of memory.

[0126] The memory 102 in the embodiment of the present invention is used to store various categories of data to support the operation of the electronic terminal 100. Examples of these data include: any executable program for operating on the electronic terminal 100, such as an operating system 1021 and an application 1022; the operating system 1021 includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks. The application 1022 may include various applications, such as a media player (MediaPlayer), a browser (Browser), etc., for implementing various application services. The method for predicting pathogenicity of germline site variations provided in the embodiment of the present invention may be included in the application 1022.

[0127] The method disclosed in the above embodiment of the present invention can be applied to the processor 101, or implemented by the processor 101. The processor 101 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 101 or the instruction in the form of software. The above processor 101 may be a general processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor 101 can implement or execute the methods, steps and logic block diagrams disclosed in the embodiment of the present invention. The general processor 101 can be a microprocessor or any conventional processor, etc. The steps of the method provided in combination with the embodiment of the present invention can be directly embodied as a hardware decoding processor to perform, or a combination of hardware and software modules in the decoding processor to perform. The software module can be located in a storage medium, which is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0128] In an exemplary embodiment, the electronic terminal 100 may be implemented by one or more application specific integrated circuits (ASIC), DSP, programmable logic device (PLD), complex programmable logic device (CPLD) to execute the aforementioned method.

[0129] The present invention also provides a computer program product, comprising a computer program, which implements the steps of the method for predicting renal function progression when executed by a processor.

[0130] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0131] Before further describing the specific embodiments of the present invention, it should be understood that the scope of protection of the present invention is not limited to the specific embodiments described below; it should also be understood that the terms used in the examples of the present invention are for describing the specific embodiments rather than for limiting the scope of protection of the present invention; in the present specification and claims, unless otherwise expressly stated herein, the singular forms "a", "an" and "the" include plural forms.

[0132] When the embodiments give numerical ranges, it should be understood that, unless otherwise specified in the present invention, both endpoints of each numerical range and any numerical value between the two endpoints can be selected. Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those generally understood by those skilled in the art. In addition to the specific methods, equipment, and materials used in the embodiments, according to the grasp of the prior art by those skilled in the art and the record of the present invention, any methods, equipment, and materials of the prior art similar or equivalent to the methods, equipment, and materials described in the embodiments of the present invention can also be used to realize the present invention.

[0133] Example

[0134] 1. Discovery Queue

[0135] The potential of urine peptide profiling as a predictive tool for predicting kidney disease progression in a general population setting was evaluated using data from the Guangzhou Nutrition and Health Study (GNHS). The GNHS is a longitudinal community-based cohort with a six-year follow-up period that included 1,664 participants. Based on estimated glomerular filtration rate (eGFR) and urine protein-to-creatinine ratio (ACR), participants were divided into:

[0136] Control group (normal renal function group, defined as eGFR ≥ 90 mL / min / 1.73 m 2 and ACR<30mg / g)

[0137] eGFR G2 group (renal function marginal group, defined as 60 mL / min / 1.73 m 2 ≤eGFR60<90mL / min / 1.73m 2 and ACR<30mg / g)

[0138] and KD group (abnormal renal function group, defined as eGFR < 60 mL / min / 1.73 m 2 or ACR ≥ 30 mg / g).

[0139] Of these, 772 participants were followed up for 6 years (n=772) to observe the progression of renal function and to evaluate the value of urinary peptides as potential biomarkers for predicting renal function progression.

[0140] 2. Exposure factors

[0141] Baseline urine peptide abundance

[0142] 3. Primary endpoint events and secondary endpoint events in the follow-up population

[0143] Primary endpoint: eGFR decline of 20% or more during follow-up (after 6 years);

[0144] Secondary endpoint event: whether the diagnostic criteria for kidney disease were met in the follow-up cohort. Specifically, participants with an ACR of 30 or above or an eGFR of less than 60 at the follow-up time point were considered to have met the secondary outcome.

[0145] 4. Urine Sample Collection

[0146] Urine samples were collected from participants during the baseline and follow-up phases of the study. Before collection, participants first cleaned their hands thoroughly with soap and water. Then, the external genital area was cleaned from front to back using mild antibacterial wipes. After the initial portion of urine was discharged, participants were asked to collect midstream urine into a sterile wide-mouth container, ensuring that the urine did not come into contact with the skin or any other surface during the process. After approximately 8 ml of midstream urine was collected, the container was immediately sealed and the outer packaging was labeled with the participant's unique identification, the date and time of sampling. Thereafter, the urine sample was centrifuged at 3000 revolutions per minute (rpm) for 15 minutes. After centrifugation, the supernatant was carefully transferred and stored in an ultra-low temperature freezer at -80°C. During the entire processing and storage process, all urine samples were handled according to uniform standard operating procedures and stored in the same containers. All demographic information that could identify the participants was removed before the samples were sent to the biobank and laboratory. During the experiment, laboratory personnel remained blinded to the specific grouping of the participants.

[0147] 5. Urine sample pretreatment

[0148] Urine samples were mixed with matrix solution containing α-cyano-4-hydroxycinnamic acid (0.3 g / L) in a ratio of 1:10. After mixing, 2 μL of the solution was applied to the target plate and allowed to fully crystallize at 37°C. To reduce bias and systematic errors, the samples were spotted in a random order, and the experimenters were blinded to the disease status of the samples. Matrix-assisted laser desorption time-of-flight mass spectrometry (MALDI-TOF MS) mass spectrometry analysis used a QuanTOF mass spectrometer, which was operated in linear positive mode and detected positively charged ions in the range of 1000 to 10,000 Da. The mass spectra were externally calibrated with calibration standards containing specific peptides to ensure that the mass deviation was controlled within 2 ppm. The ion source parameter settings included: source voltage 20 kV, detector voltage 0.48 kV, laser pulse energy 4.8 μJ, pulse frequency 3000 Hz, focal mass 5000 Da, scanning speed 1 mm / sec, and 800 pulses averaged per spectrum. Finally, the mass spectra were acquired and processed using QuanTOF Viewer software.

[0149] 6. MALDI-TOF (matrix-assisted laser desorption time-of-flight mass spectrometry) data analysis

[0150] The method is similar to the invention patent with application number 202210926734.7, and is briefly described as follows: The original mass spectrometry file is first imported into the R software and processed using the "MALDIquantForeign" R package (version 0.12). Subsequently, the mass spectrometry data is processed using the "MALDIquant" R package (version 1.19.3) in the following steps:

[0151] ① Variance stabilization: Reduce possible variance-mean dependence through square root transformation.

[0152] ② Smoothing and baseline correction: The spectrum was smoothed using a 21-point Savitzky Golay filter, and the baseline was corrected using the SNIP algorithm.

[0153] ③Peak intensity value calibration: In order to facilitate comparison and reduce batch effects, the peak intensity values ​​were equalized by total ion current calibration (TIC).

[0154] ④ Spectrum alignment and peak detection: The warp algorithm was used to recalibrate the mass value, and peak detection was performed with a threshold of signal-to-noise ratio (SNR) greater than 2.

[0155] ⑤Peak merging and feature matrix generation: merge peaks with similar mass values.

[0156] After the above processing, a feature matrix suitable for subsequent statistical analysis is generated.

[0157] 7. Characteristic peak identification

[0158] The urine peptide peaks that can be used as important markers of renal function were screened by the following method:

[0159] The method is similar to the invention patent with application number 202210926734.7, and the steps are as follows Figure 1 As shown in D, when screening key peaks for the detection of abnormal kidney function, the baseline urine peptide data of the community population were grouped and compared: control group and kidney disease group samples. By comparing the peak differences between the two groups, it was subsequently validated in validation dataset 1 and validation dataset 2. Three different machine learning techniques were used to screen peaks related to kidney disease: partial least squares discriminant analysis (PLS-DA), least absolute shrinkage and selection operator (LASSO), and recursive feature elimination with cross validation (RFECV). For PLS-DA, the VIP score of the peak was calculated, and the top 10 peaks with VIP greater than 1 were selected as disease features. Using the LASSO method, 80% of the samples were randomly selected, repeated 200 times, and peaks with a frequency of more than 50% were considered as kidney disease features. Through RFECV, the top 10 peaks were selected based on the accuracy of cross validation. After combining these characteristic peaks and verifying their true presence on the spectra, they were identified as important urine peptide pattern components for distinguishing kidney disease (KD) from healthy controls. This approach enables the identification of statistically significant urine peptide patterns that can serve as biomarkers for distinguishing healthy and disease states.

[0160] Principal coordinate analysis (PCoA) was then used to explore whether changes in urine peptide patterns showed group-specific trends. Based on the abundance of urine peaks, a diversity difference matrix was calculated using three distance metrics: Bray-Curtis distance, binomial distance, Euclidean distance, and Jaccard distance. Permutational multivariate analysis of variance (PERMANOVA) was performed after 1000 permutations to evaluate the differences in urine peptide model diversity between groups in the PCoA plot.

[0161] Using the above machine learning algorithm, 15 urine polypeptide peaks that changed significantly under renal disease conditions were screened out. The 15 urine polypeptide peaks with mass-to-charge ratios can be used as important markers of renal function, as follows: "m / z 1378.8", "m / z1725.95", "m / z1902.7", "m / z 1933.8", "m / z 2234.26", "m / z 2752.42", "m / z 1013.59", "m / z 1235.53", "m / z1911.76", "m / z 2039.86", "m / z 1949.9", "m / z 1869.59", "m / z2086.71", "m / z 1641.2", and "m / z3321.3".

[0162] 8. Validation in two external independent validation populations

[0163] The independent validation population 1 consisted of 528 newly recruited participants in the cross-sectional study of the Guangzhou Nutrition and Health Study (GNHS) from 2017 to 2020. The population was divided into three groups based on estimated glomerular filtration rate (eGFR) and urine albumin / creatinine ratio (ACR):

[0164] The control group (normal renal function) included 96 patients with eGFR ≥ 90 mL / min / 1.73 m 2 Participants with ACR < 30 mg / g,

[0165] eGFR G2 group (renal function in the marginal zone) included 332 patients with 60 ≤ eGFR < 90 mL / min / 1.73 m 2 Participants with ACR < 30 mg / g,

[0166] The KD group (abnormal renal function) included 100 patients with eGFR < 60 mL / min / 1.73 m 2 or participants with ACR ≥ 30 mg / g.

[0167] Inclusion and exclusion criteria and procedures for collecting urine and blood samples, performing laboratory tests, and urine peptide analysis remained consistent with the discovery cohort.

[0168] Independent validation population 2 was a cross-sectional population from a hospital. Between April 2019 and January 2020, patients with confirmed chronic kidney disease (CKD) and healthy volunteers were recruited from Zhujiang Hospital, Nanfang Hospital, and Guangdong Provincial People's Hospital. The study was approved by the Ethics Committee of Zhujiang Hospital of Southern Medical University (approval number 2020-KY-016-01) and registered with the Chinese Clinical Trial Registry with registration number ChiCTR2000033229 (www.chictr.org.cn). 194 volunteers with abnormal renal function confirmed by renal biopsy pathology in the Department of Nephrology were included in the CKD group. In addition, 48 volunteers with normal renal function (eGFR ≥ 90 mL / min / 1.73 m 2 The CKD group was matched with the control group (ACR < 30 mg / g) to ensure age and sex matching. The process of collecting urine and blood samples, laboratory tests, and urine peptide analysis followed the same process as the discovery cohort.

[0169] In the three cohorts (discovery cohort and two other external independent validation cohorts), significant changes in urine peptide models were found between groups, and there was a certain trend between groups ( Figure 1 AC, PCOA dimensionality reduction analysis was performed, and the groups were divided into normal renal function group (Control), marginal renal function group (eGFR G2) and abnormal renal function group (KD). The verification cohort 2 was CKD patients diagnosed by renal biopsy and healthy normal renal function group (Control). The inter-group differences of these 15 key urinary peptide peaks were observed in the three cohorts, and it was found that the inter-group differences of key urinary peptide peaks could also be reproduced in two independent external populations, and some urine peptide peaks had already undergone significant changes in the eGFR G2 subgroup (renal function between control and KD) ( Figure 1 Middle E).

[0170] 9. Construction of urine peptide index based on 15 urine peptide peaks

[0171] ① Random Forest Algorithm

[0172] In the urine peptide data of the community population baseline cohort, a classification model was established using a machine learning algorithm based on the selected 15 urine peptide peaks. Taking the random forest model as an example, in the discovery cohort, a training data set was constructed based on the relative peak intensity of 15 urine peptide peaks in healthy renal function samples and kidney disease samples. Through five rounds of repeated ten-fold cross-validation, the average predicted probability of kidney disease in the training population was calculated and used as the total urine peptide index of the population.

[0173] Subsequently, the random forest (RF) model was applied to the eGFR G2 subgroup (i.e., renal function between normal and abnormal, defined as 60≤eGFR<90 and ACR<30) in the community population and two additional independent cohorts to predict the probability of an individual developing kidney disease, and these prediction results were used as the urine peptide index for the corresponding individuals.

[0174] AUC is used to evaluate the performance of the model. The results are as follows Figure 2 As shown, in the baseline population of the discovery cohort, the random forest classification model based on 15 urine peptides to distinguish between normal renal function (Control group) and abnormal renal function (KD group) had an area under the ROC curve (AUROC) of 77% in 10-fold cross-validation. The model was also validated using two independent cohorts. For external validation set 1 (a cross-sectional cohort based on the GNHS community population), the AUROC was 81%; while the AUROC for external validation set 2 (a cohort of CKD patients confirmed by biopsy) exceeded 91%. We also applied multiple machine learning algorithms and extended the model to two independent validation cohorts, and its performance was consistent with the random forest model. This shows that the model based on 15 peptide peaks is robust and reliable in distinguishing healthy individuals from individuals with renal insufficiency.

[0175] ② Calculation method

[0176]

[0177] UPI i represents the urine peptide index of participant i,

[0178] n is the number of key urine peptides identified by the above screening process (7. Characteristic peak identification) (n=15).

[0179] Intensity ij Represents the relative peak intensity of the jth urine peptide in participant i.

[0180] OR j represents the odds ratio (OR) of the j-th urine peptide obtained by univariate logistic regression analysis, which is constructed using the healthy renal function group and the renal disease group in the baseline population of the discovery cohort, with the relative peak intensity of each urine peptide fragment of an individual as the independent variable and whether the individual has abnormal renal function as the dependent variable.

[0181] When the jth urine polypeptide acts as a protective factor for renal function, that is, log 10 (OR j ) is a negative value; on the contrary, when it is used as a risk factor for renal function, log10 (OR j ) is a positive value.

[0182] Take the following data as an example to show how to use the formula:

[0183] Participant

[0184] Number of urine peptide fragments included in the analysis n = 15

[0185] According to MALDI-TOF, the relative peak intensity of the first urine peptide was obtained as Intensityi1=0.5, and the odds ratio OR1=2;

[0186] According to MALDI-TOF, the relative peak intensity of the second urine peptide was obtained, Intensityi2 = 0.6, and the odds ratio OR2 = 0.5;

[0187] According to MALDI-TOF, the relative peak intensity of the third urine peptide was obtained as Intensityi3=0.7, and the odds ratio OR3=1.2;

[0188] The relative peak intensities of the remaining urine peptide fragments were assumed to be zero.

[0189] The calculation steps are as follows:

[0190] UPI i =(0.5×log 10 (2))+(0.6×log 10 (0.5))+(0.7×log 10 (1.2))

[0191] =0.1505+(-0.1806)+0.055426

[0192] =0.025326

[0193] The weighted sum of all fragments of participant i is accumulated to obtain the urine peptide index UPI of participant i i .

[0194] 10. Association of urine peptide index with primary / secondary endpoint events

[0195] Univariate and multivariate-adjusted logistic regression analyses were performed to assess the ability of the urine peptide index as a risk marker for kidney function decline at baseline, i.e., to predict a ≥20% decline in the estimated glomerular filtration rate (eGFR) at follow-up (primary endpoint) and to predict whether or not a kidney disease diagnosis criteria was reached at follow-up (secondary endpoint).

[0196] Logistic regression was used to estimate the probability of an event, given a data set of independent variables. Urinary peptide index was used as the independent variable and the occurrence of a primary or secondary endpoint event was used as the dependent variable. Subsequently, confounding factors were included in the model for correction.

[0197] Model 1 (model-1) is an unadjusted logistic regression model.

[0198] Model-2 was adjusted for age (continuous variable), sex (female / male), BMI (continuous variable), hypertension (yes / no), and diabetes (yes / no). Hypertension was defined as systolic blood pressure ≥140 mmHg and / or diastolic blood pressure ≥90 mmHg, or a history of hypertension diagnosis, or the use of antihypertensive drugs. Diabetes was defined as fasting blood glucose ≥7 mmol / L, or glycosylated hemoglobin HbA1c ≥6.5%, or a history of diabetes diagnosis, or the use of hypoglycemic drugs.

[0199] Model 3 (model-3) further adjusted the baseline eGFR (continuous variable) and urine albumin / creatinine ratio (ACR, continuous variable) on the basis of model 2, that is, correction was made by incorporating the two variables of eGFR and ACR into the regression model. Specifically, in the logistic regression model, eGFR and ACR were included in the model as additional independent variables (continuous variables) to control their potential effects on primary or secondary endpoint events. The purpose of this step is to reduce the interference of confounding factors so that the model can more accurately estimate the association between urine polypeptide index (UPI) and endpoint events. For example: Assuming that the probability of the occurrence of the primary endpoint event is to be estimated, the formula of the logistic regression model is as follows:

[0200] logit(P) = β0+β1×UPI+β2×ageβ3×sex+β4×BMI+β5×hypertension+β6×diabetes+β7×eGFR+β8×ACR

[0201] in:

[0202] P is the probability of the primary endpoint occurring

[0203] β0 is the intercept term

[0204] β1,β2,...,β8 are the regression coefficients of each independent variable

[0205] UPI is the urine polypeptide index

[0206] Age, sex, BMI, hypertension, and diabetes were adjustment variables in Model 2

[0207] eGFR is the estimated glomerular filtration rate at baseline

[0208] ACR is the urine albumin / creatinine ratio

[0209] By incorporating eGFR and ACR as additional independent variables, the model was able to more comprehensively correct for the effects of these biomarkers, making the analysis results more accurate.

[0210] Results:

[0211] Results were expressed as odds ratios (ORs) and 95% confidence intervals (95% CIs). UPI was used as both a continuous variable and a tertile ordinal categorical variable as an independent variable in the logistic regression model. The significance of the trend was assessed by replacing each tertile of the UPI categorical variable with values ​​1-3 as a continuous independent variable (p-trend analysis).

[0212] ① Results of the random forest algorithm

[0213] Analysis methods: In the discovery cohort, 772 people were followed up effectively after 6 years (including urine collection for urine peptide detection and renal function-related tests). A logistic prospective analysis was performed with the urine peptide index constructed by the random forest algorithm of Part 10 with 15 peaks as the independent variable and the occurrence of primary / secondary endpoint events as the dependent variable.

[0214] The association between the baseline urine peptide index and the primary endpoint events constructed using the random forest RF model based on 15 urine peptide peaks is as follows Figure 3 As shown in , among 772 people, 65 reached the primary endpoint (eGFR decreased ≥ 20%). The association between the baseline urine peptide index constructed by the random forest RF model based on 15 urine peptide peaks and the secondary endpoint events is shown in Figure 4 As shown, among 772 people, 157 reached the secondary endpoint event (met the KD diagnostic criteria during follow-up).

[0215] ②Results of the formula calculation method

[0216] Analysis methods: In the discovery cohort, 772 people were followed up effectively after 6 years (including urine collection for urine peptide detection and renal function-related tests). The urine peptide index constructed by the calculation method of 15 peaks according to formula ① was used as the independent variable, and the occurrence of primary / secondary endpoint events was used as the dependent variable for logistic prospective analysis. Subsequently, confounding factors were included in the model for correction. The details of the confounding factors for model correction are described in Section 10. Model 1 (model-1) in the following table is an unadjusted logistic regression model. Model 2 (model-2) was adjusted according to age (continuous variable), sex (female / male), BMI (continuous variable), hypertension (yes / no) and diabetes (yes / no). Model 3 (model-3) further adjusted the baseline eGFR (continuous variable) and urine albumin / creatinine ratio (ACR, continuous variable). The association results of the baseline urine peptide index constructed by the formula based on 15 urine peptide peaks with primary endpoint events and secondary endpoint events are shown in Table 1 and Table 2, respectively.

[0217] Table 1 Correlation results between baseline urine peptide index and primary endpoint events based on the formula constructed based on 15 urine peptide peaks

[0218]

[0219] Table 2 Correlation results between baseline urine peptide index constructed by formula based on 15 urine peptide peaks and secondary endpoint events

[0220]

[0221] In Tables 1 and 2 above, Model is listed as the model under different correction confounding factors; urine polypeptide index (UDI) is used as the independent variable in the logistic regression equation in the form of a continuous variable, and the OR (taking Model 3 in Table 1 as an example) is expressed as the probability of the occurrence of the primary endpoint event increases by 53% (OR = 1.53, 95% confidence interval 1.17-2.01, p = 0.002) for each SD (i.e. standard deviation) increase in the independent variable (urine polypeptide index); Figure 3 and Figure 4 The results in are similar.

[0222] The above examples are for the purpose of illustrating the embodiments disclosed by the present invention and are not to be construed as limiting the present invention. In addition, the various modifications listed herein and the variations of the methods in the invention are obvious to those skilled in the art without departing from the scope and spirit of the present invention. Although the present invention has been specifically described in conjunction with various specific preferred embodiments of the present invention, it should be understood that the present invention should not be limited to these specific embodiments. In fact, various modifications as described above that are obvious to those skilled in the art to obtain the invention should be included within the scope of the present invention.

Claims

1. Use of urine polypeptides in the preparation of early detection products for renal function progression, characterized in that: The mass-to-charge ratio of the urine polypeptide is any one or more of the following: m / z 1378.8, m / z 1725.95, m / z 1902.7, m / z 1933.8, m / z2234.26, m / z 2752.42, m / z 1013.59, m / z 1235.53, m / z 1911.76, m / z 2039.86, m / z1949.9, m / z 1869.59, m / z 2086.71, m / z 1641.2, m / z 3321.3, and the allowable deviation range of the mass-to-charge ratio is 500 ppm.

2. The use according to claim 1, characterized in that The urine polypeptides include any one or more of the following characteristics: 1) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1378.8 is shown in SEQ ID NO.1; 2) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1725.95 is shown in SEQ ID NO.2; 3) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1902.7 is shown in SEQ ID NO.3; 4) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1933.8 is shown in SEQ ID NO.4; 5) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2234.26 is shown in SEQ ID NO.5; 6) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2752.42 is shown in SEQ ID NO.6, 7) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1013.59 is shown in SEQ ID NO.7, 8) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1235.53 is shown in SEQ ID NO.8, 9) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1911.76 is shown in SEQ ID NO.9, 10) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2039.86 is shown in SEQ ID NO.10, 11) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1949.9 is shown in SEQ ID NO.11, 12) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1869.59 is shown in SEQ ID NO.12, 13) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 2086.71 is shown in SEQ ID NO.13, 14) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 1641.2 is shown in SEQ ID NO.14, 15) The amino acid sequence of the urine polypeptide with a mass-to-charge ratio of m / z 3321.3 is shown in SEQ ID NO.

15.

3. The use according to claim 1, characterized in that The early detection product for renal function progression is an early detection product for renal disease progression; preferably, the renal disease is chronic kidney disease.

4. A product for early detection of renal function progression, characterized in that: The early detection product includes any one or more of the urine polypeptides for use according to claim 1, or contains antibodies that can specifically bind to the urine polypeptides.

5. A method for constructing an early detection model for renal function progression, characterized in that: The construction method comprises the following steps: S1, obtaining relative peak intensity data of urine polypeptides used in the use of claim 1 or 2 for people with normal renal function and people with abnormal renal function, and using the obtained relative peak intensity data of each urine polypeptide as a training data set to construct a machine learning model, and directly or indirectly obtaining a urine polypeptide index of the sample to be tested according to the constructed machine learning model; S2, using the urine polypeptide index obtained in S1 as the independent variable, and whether the primary or secondary endpoint event occurs as the dependent variable to construct a logistic regression model, the logistic regression model is used for the early detection of renal function progression, the primary endpoint event refers to a significant decrease in the renal disease index of the sample to be tested when the predetermined time is reached; the secondary endpoint event refers to the sample to be tested reaching the renal disease diagnostic criteria when the predetermined time is reached.

6. The construction method according to claim 5, characterized in that: The relative peak intensity data of urine polypeptides of people with normal renal function and people with abnormal renal function described in S1 are data obtained by MALDI-TOF-MS detection; or, the machine learning model described in S1 is selected from any one or more of the following: support vector machine, naive Bayes, random forest, K-nearest neighbor, gradient boosting decision tree, decision tree, adaptive boosting or logistic regression.

7. The construction method according to claim 5, characterized in that: The machine learning model described in S1 is a random forest model, which is constructed in the following manner: obtaining relative peak intensity data of urine polypeptides used in the use described in claim 1 or 2 for people with normal renal function and people with abnormal renal function, and using the obtained relative peak intensity data of each urine polypeptide as a training data set for model training, and performing an accuracy test, and obtaining the optimal model as the urine polypeptide index calculation model based on the accuracy test results.

8. The construction method according to claim 7, characterized in that: The method for testing the accuracy of the model includes independent validation or cross validation; preferably, the cross validation is selected from k-fold cross validation; preferably, the k-fold cross validation is ten-fold cross validation.

9. The construction method according to claim 5, characterized in that: The machine learning model in S1 is a logistic regression model, preferably, a univariate logistic regression model; more preferably, when the machine learning model in S1 is a logistic regression model, S1 comprises the following steps: S11, model construction: obtaining relative peak intensity data of urine polypeptides used in the use of claim 1 or 2 for people with normal renal function and people with abnormal renal function, using the relative peak intensity data of each urine polypeptide obtained as an independent variable, and constructing a logistic regression model with whether or not there is abnormal renal function as a dependent variable; S12, calculating the urine polypeptide index, obtaining the OR value of each urine polypeptide according to the logistic regression model constructed in S11, substituting the OR value into formula ① to calculate the urine polypeptide index of the sample to be tested; Among them, UPIi represents the urine polypeptide index of the sample i to be tested, n is the number of urine peptides, n=15, intensity ij represents the relative peak intensity of the jth urine peptide in the sample i to be tested. ORj represents the odds ratio of the jth urine peptide obtained by the logistic regression model.

10. The construction method according to claim 5, characterized in that: The logistic regression model in S2 is a univariate logistic regression model or a multivariate logistic regression model; preferably, the independent variables in the multivariate logistic regression model also include any one or more of the following confounding variables: age, gender, BMI, whether the patient has hypertension, whether the patient has diabetes, eGFR, ACR.

11. The construction method according to claim 5, characterized in that: The renal disease index in S2 is selected from eGFR and / or ACR; preferably, a significant decrease in the renal disease index in S2 means that the eGFR of the sample to be tested decreases by more than 15% relative to the sampling time point of the data obtained in S1.

12. A model for early detection of renal function progression obtained by the model construction method according to any one of claims 5 to 11.

13. A method for early detection of renal function progression, characterized in that: The method comprises the following steps: 1) Calculating the urine polypeptide index: inputting the relative peak intensity data of the urine polypeptide of the sample to be tested into the machine learning model obtained in step S1 of the construction method described in any one of claims 5 to 11, and directly or indirectly obtaining the urine polypeptide index of the sample to be tested, wherein the urine polypeptide is the urine polypeptide in the use described in claim 1 or 2; 2) Early detection of renal function progression and output of results: the urine polypeptide index obtained in step 1) is input into the logistic regression model obtained in step S2 of the construction method described in any one of claims 5 to 11, and the early detection model for renal function progression outputs the result of whether a major endpoint event or a secondary endpoint event occurs. The major endpoint event refers to a significant decrease in the renal disease index of the sample to be tested when a predetermined time is reached; the secondary endpoint event refers to a sample to be tested reaching the renal disease diagnostic criteria when a predetermined time is reached.

14. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method for early detection of renal function progression as claimed in claim 13 is implemented.

15. A computer processing device comprising a processor and the computer-readable storage medium according to claim 14, characterized in that: The processor executes the computer program on the computer-readable storage medium to implement the steps of the method for early detection of renal function progression according to claim 13.

16. An electronic terminal, characterized in that: include: Processor, memory and communicator; The memory is used to store computer programs, the communicator is used to communicate with an external device, and the processor is used to execute the computer program stored in the memory so that the terminal executes the early detection method for renal function progression as described in claim 13.

17. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for early detection of renal function progression according to claim 13 are implemented.

Images