A combination of oligosaccharide markers for differentiating pathological subtypes of multiple myeloma and its application.

Abstract

This invention discloses a combination of oligosaccharide biomarkers for identifying pathological subtypes of multiple myeloma and its application. The combination comprises NGA2F, NG1A2F, NG1A2FB, NA2B, NA2F, NA2FB, NA3, NA3Fb, NA4, and NA4Fb. This invention is the first to discover and validate a biomarker combination consisting of 10 specific oligosaccharide chains, the abundance of which is closely related to different pathological subtypes of MM. Based on this discovery, this invention provides a novel application of this biomarker combination in the preparation of subtyping detection kits and further establishes a method for constructing a high-precision subtyping prediction model using machine learning algorithms. This model can achieve rapid and accurate MM subtyping identification based on the oligosaccharide chain map, with an overall concordance rate exceeding 94% on the validation set. This invention provides a novel, non-invasive solution for the accurate subtyping of multiple myeloma and has significant clinical value.

Description

Technical Field

[0001] This invention belongs to the field of molecular biomedical technology, specifically relating to a combination of oligosaccharide chain markers for identifying pathological subtypes of multiple myeloma and their applications. Background Technology

[0002] Multiple myeloma (MM) is a malignant proliferative tumor originating from plasma cells. It is the second most common hematologic malignancy, accounting for approximately 10% of all hematologic malignancies. The disease primarily affects the elderly, and although treatment methods have improved in recent years, it remains incurable, posing a serious threat to patients' lives and health.

[0003] Accurate pathological subtyping is crucial in the clinical management of multiple myeloma (MM). MM can be classified into different subtypes based on the type of monoclonal immunoglobulins secreted, including IgG, IgA, light chain, IgD, and IgM. Different subtypes exhibit significant differences in disease biological behavior, treatment options, and prognosis. Therefore, precise subtyping allows clinicians to gain a deeper understanding of the individual characteristics of the disease, enabling the development of more targeted and individualized treatment strategies. This is of decisive significance for improving treatment efficacy, reducing unnecessary drug side effects, improving patients' quality of life, and prolonging survival.

[0004] Currently, clinical classification of multiple myeloma (MM) primarily relies on serum protein electrophoresis and immunofixation electrophoresis. However, these conventional methods have limitations, particularly in terms of detection sensitivity, which may lead to missed detections of trace amounts of M protein or inaccurate classification, failing to fully meet the increasingly demanding requirements for classification in the era of precision medicine. Therefore, exploring and discovering novel biomarkers that can predict MM patient classification with high sensitivity and specificity has become a key technical problem that urgently needs to be solved in current clinical practice.

[0005] In recent years, scientific research has shown that the occurrence and development of multiple myeloma (MM) are closely related to abnormal protein glycosylation. MM patients produce a large number of abnormal glycoproteins, and the oligosaccharide chains attached to these glycoproteins undergo significant changes in both structure and number. These changes in glycosylation are considered an important event in tumor development and progression. However, current research mainly focuses on the association between glycosylation and the occurrence or prognosis of MM. In-depth and systematic research on its potential link with specific pathological subtypes of MM is lacking, and a reliable method for predicting clinical subtypes has not yet been established.

[0006] In summary, there is an urgent need in this field for a novel and efficient solution to overcome the shortcomings of existing MM genotyping techniques in terms of sensitivity. Given the significant alterations of oligosaccharide chains in MM, exploring their association with MM genotyping and developing a non-invasive, highly sensitive genotyping prediction tool based on oligosaccharide chains has significant clinical value and broad application prospects. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a combination of oligosaccharide chain biomarkers for identifying pathological subtypes of multiple myeloma and its application. By discovering a specific set of oligosaccharide chain biomarkers for the first time and constructing a high-precision prediction model, rapid, non-invasive, and highly accurate identification of different pathological subtypes of multiple myeloma is achieved, providing a key basis for individualized clinical treatment.

[0008] This invention is achieved through the following technical solution:

[0009] An oligosaccharide chain biomarker combination for identifying pathological subtypes of multiple myeloma, the combination consisting of 10 specific oligosaccharide chains, namely NGA2F, NG1A2F, NG1A2FB, NA2B, NA2F, NA2FB, NA3, NA3Fb, NA4, and NA4Fb.

[0010] A method for constructing a predictive model for pathological subtyping of multiple myeloma includes the following steps:

[0011] Step 1) Obtain a training sample set, which includes biological samples of multiple myeloma patients with known pathological subtypes and their corresponding subtype labels;

[0012] Step 2) Detect the abundance of each oligosaccharide chain in the above-mentioned oligosaccharide chain marker combination for each sample in the training sample set, and obtain oligosaccharide chain abundance data;

[0013] Step 3) Using the oligosaccharide chain abundance data as the independent variable and the genotype label as the dependent variable, a machine learning algorithm is used to train the model to obtain the prediction model.

[0014] Preferably, the biological sample in step 1) is blood, serum, or plasma from venous or peripheral blood of a multiple myeloma patient.

[0015] Preferably, the machine learning algorithm in step 3) is one or more of logistic regression, random forest, support vector machine, neural network, and XGBoost algorithm.

[0016] Preferably, the pathological subtypes include IgG type, IgA type, light chain type, IgD type and IgM type.

[0017] A pathological subtyping prediction model for multiple myeloma obtained through the above construction method.

[0018] A computer-readable storage medium storing the above-described prediction model.

[0019] A pathological classification prediction system for multiple myeloma, comprising:

[0020] The data acquisition module is used to acquire the abundance information of the above-mentioned oligosaccharide chain marker combination in the sample to be tested;

[0021] The fractal prediction module stores the aforementioned prediction model.

[0022] The results output module is configured to output the pathological classification prediction results calculated by the prediction model based on the abundance information.

[0023] The above-mentioned oligosaccharide chain marker combination is used in the preparation of a kit to assist in the identification of pathological subtypes of multiple myeloma.

[0024] A kit for assisting in the identification of pathological subtypes of multiple myeloma, comprising reagents for detecting the aforementioned combination of oligosaccharide chain markers.

[0025] The beneficial effects of this invention are as follows:

[0026] (1) This invention is the first to discover and verify a group of biomarkers consisting of 10 specific oligosaccharide chains, whose abundance information has a highly specific correlation with different pathological subtypes of MM (including IgG type, IgA type, light chain type, IgD type and IgM type). This provides an unprecedented and novel biomarker system for the field of MM subtyping, breaking through the limitations of traditional immunoglobulin-dependent detection.

[0027] (2) This invention combines specific oligosaccharide chain markers with machine learning algorithms (such as support vector machines) to construct a multiple myeloma classification prediction model (Glycan Pathological Classification System, GPCS), which exhibits excellent predictive performance. Experimental results show that the model achieves a total pathological classification accuracy of over 94% in both the training and validation sets, and a classification accuracy of up to 100% for the clinically rare IgD type MM. This high accuracy and high sensitivity effectively overcome the potential sensitivity limitations of existing techniques such as immunofixation electrophoresis.

[0028] (3) The test sample relied upon by this invention is blood (blood, serum and plasma in venous blood or peripheral blood), which is easy to obtain, non-invasive, and has good patient compliance. Compared with invasive examinations such as bone marrow aspiration, the method provided by this invention is easier to repeat sampling and dynamic monitoring, creating favorable conditions for promotion in primary hospitals and the realization of long-term patient condition management.

[0029] (4) Through the classification prediction method provided by the present invention, clinicians can obtain the patient's pathological classification information more quickly and accurately, thereby providing a strong objective basis for formulating individualized treatment plans (such as selecting targeted drugs for specific immunoglobulin types) and assessing prognosis, and ultimately improving the patient's treatment effect and quality of life. Attached Figure Description

[0030] Figure 1 The oligosaccharide chain profiles of MM patients with different subtypes in Example 1 are as follows: A is IgG type; B is IgA type; C is light chain type; D is IgM type; E is IgD type. Detailed Implementation

[0031] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0032] Unless otherwise specified, the technical means used in the following embodiments are all conventional means well known to those skilled in the art, and the experimental methods without specific conditions are all conventional methods in the art.

[0033] Unless otherwise specified, all materials and reagents used in the following examples are commercially available.

[0034] Example 1

[0035] 1. Test Sample

[0036] This embodiment collected blood samples from 695 MM patients with different subtypes as a dataset, including 280 IgG subtype samples, 140 IgA subtype samples, 140 light chain subtype samples, 70 IgM subtype samples, and 65 IgD subtype samples. All samples were obtained from Nanjing University Medical School Affiliated Gulou Hospital, ethics number 2023-248-01.

[0037] The dataset was randomly divided into a training set and a validation set in a 7:3 ratio, with 500 cases in the training set and 195 cases in the validation set. In the training set, there were 210 IgG samples, 98 IgA samples, 98 light chain samples, 49 IgM samples, and 45 IgD samples. In the validation set, there were 70 IgG samples, 42 IgA samples, 42 light chain samples, 21 IgM samples, and 20 IgD samples.

[0038] 2. Instruments and equipment

[0039] Capillary electrophoresis analyzer, PCR instrument and centrifuge.

[0040] 3. Test reagents

[0041] Reagent A: 5 mM NH4HCO3 added to 1% SDS solution;

[0042] Reagent B: Add 2 U / μL of exoglycoside exonuclease solution to 1% NP-40;

[0043] Reagent C: Add 2 U / μL of glycoside endonuclease solution to 100 mM, pH 5 NH4AC;

[0044] Reagent D: ddH2O;

[0045] Reagent E: A solution prepared by mixing 5 mM fluorescent labeling solution (trisodium 8-aminopyrene-1,3,6-trisulfonic acid) with DMSO solution (organic reducing agent NaBH3CN concentration of 1 M).

[0046] 4. Oligosaccharide chain pattern detection and collection

[0047] (1) Release of oligosaccharide chains

[0048] Add 3 μL of reagent A to 5 μL of sample, heat at 95℃ for 5 min to denature, cool to room temperature, add 3 μL of reagent B and 4 μL of reagent C, react at 37℃ for 4 h, and then add 80 μL of reagent D.

[0049] (2) Marking of oligosaccharide chains

[0050] Take 10 μL of the sample solution from step (1), dry it at 70℃ for 30 min, then add 3 μL of reagent E, react at 90℃ for 2 h, and finally add 80 μL of reagent D to terminate the reaction.

[0051] (3) Detection of oligosaccharide chains and acquisition of spectra

[0052] Take 10 μL of the oligosaccharide chain sample prepared in step (2), place it in an ABI-specific 96-well plate, and detect it using an ABI3500 sequencer to obtain the oligosaccharide chain map.

[0053] like Figure 1As shown, the relative amounts of 10 specific oligosaccharide chains in each blood sample were obtained through protein denaturation, glycosidase treatment, fluorescent labeling, oligosaccharide chain mapping detection, and data acquisition. These 10 specific oligosaccharide chains are: NGA2F (non-galactosyl α-1,6 core fucosylated biantennary oligosaccharide chain), NG1A2F (monobranched galactosyl α-1,6 core fucosylated biantennary oligosaccharide chain), NG1A2FB (monobranched galactosyl α-1,6 core fucosylated biantennary oligosaccharide chain), N... A2B (galactosylated bipolar oligosaccharide chain), NA2F (galactosylated α-1,6-core fucosylated bipolar oligosaccharide chain), NA2FB (galactosylated α-1,6-core fucosylated bipolar oligosaccharide chain), NA3 (galactosylated tripolar oligosaccharide chain), NA3Fb (galactosylated α-1,3-branched fucosylated tripolar oligosaccharide chain), NA4 (galactosylated tetrapolar oligosaccharide chain), NA4Fb (galactosylated α-1,3-branched fucosylated tetrapolar oligosaccharide chain).

[0054] 5. Data Analysis and Processing

[0055] Depend on Figure 1 It can be seen that the oligosaccharide chain profiles of patients with different subtypes of MM are significantly different, indicating that oligosaccharide chains have the potential to distinguish between different subtypes.

[0056] (1) Screening of characteristic oligosaccharide chains

[0057] Oligosaccharide chain data among different subtypes of MM patients in the training set (500 cases) were analyzed. Comparative analysis was performed on 10 oligosaccharide chain data from 500 patients with different subtypes. Oligosaccharide chains with statistically significant p-values ​​less than 0.01 were selected as biomarkers for inclusion in the predictive model, as shown in Table 1 below.

[0058] Table 1. Comparative analysis of oligosaccharide chains in MM patients of different subtypes in the training set (mean ± standard deviation)

[0059]

[0060] (2) Constructing a classification model

[0061] Based on the selected 10 oligosaccharide chains, a classification model GPCS was constructed using the Support Vector Machine (SVM) algorithm in the training set. During model construction, the clinically determined subtype was used as the dependent variable, and the oligosaccharide chain data was used as the independent variable. The model parameters kernel, C, gamma, and random state were optimized. The final determined parameters were kernel='rbf', C=1, gamma=0.15, and random state=2024. The probability values ​​output by the model were used as predicted values, with the highest probability value being the actual result. The performance of the prediction model was validated using validation set data, as shown in Tables 2 and 3 below.

[0062] Table 2. Concordance rate between GPCS model and clinical classification results in the training set.

[0063]

[0064] As shown in Table 2, the overall concordance rate between the GPCS model in the training set and the clinical classification results reached 95.20%, with the concordance rate for IgG type being 95.71%, IgA type being 93.88%, light chain type being 94.90%, IgM type being 91.84%, and IgD type being 100.00%.

[0065] Table 3. Concordance rate between GPCS model and clinical outcomes in the validation set.

[0066]

[0067] As shown in Table 3, the overall concordance rate between the GPCS model and the clinical classification results in the validation set reached 94.87%, with the concordance rate of IgG type being 94.29%, IgA type being 92.86%, light chain type being 95.24%, IgM type being 95.24%, and IgD type being 100.00%.

[0068] The experimental results of this embodiment show that the MM pathological subtyping prediction model GPCS based on oligosaccharide chains has a concordance rate of 95.20% with clinical results in the 500 training data set and a concordance rate of 94.87% with clinical results in the 195 independent validation data set, demonstrating good performance.

[0069] The embodiments described above are only some, not all, of the embodiments of the present invention. The detailed description of the embodiments of the present invention is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments. The scope of protection of the present invention is determined by the scope claimed in the claims. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

Claims

1. A method for constructing a predictive model for pathological subtyping of multiple myeloma, characterized in that, The pathological classification includes IgG type, IgA type, light chain type, IgD type, and IgM type; the construction method includes the following steps: Step 1) Obtain a training sample set, which includes biological samples of multiple myeloma patients with known pathological subtypes and their corresponding subtype labels; Step 2) Detect the abundance of each oligosaccharide chain in the oligosaccharide chain marker combination of each sample in the training sample set to obtain oligosaccharide chain abundance data; The oligosaccharide chain marker combination consists of 10 specific oligosaccharide chains: NGA2F, NG1A2F, NG1A2FB, NA2B, NA2F, NA2FB, NA3, NA3Fb, NA4, and NA4Fb. Step 3) Using the oligosaccharide chain abundance data as the independent variable and the genotype label as the dependent variable, the support vector machine algorithm is used to train the model to obtain the prediction model.

2. The method for constructing a predictive model for pathological subtyping of multiple myeloma according to claim 1, characterized in that, Step 1) The biological samples are blood, serum and plasma from venous blood or peripheral blood of patients with multiple myeloma.

3. A pathological subtyping prediction model for multiple myeloma obtained by the construction method described in claim 1 or 2.

4. A computer-readable storage medium storing the prediction model as described in claim 3.

5. A pathological classification prediction system for multiple myeloma, characterized in that, The pathological classification includes IgG type, IgA type, light chain type, IgD type, and IgM type; the system includes: The data acquisition module is used to acquire abundance information of oligosaccharide chain marker combinations in the sample to be tested; The oligosaccharide chain marker combination consists of 10 specific oligosaccharide chains: NGA2F, NG1A2F, NG1A2FB, NA2B, NA2F, NA2FB, NA3, NA3Fb, NA4, and NA4Fb. The fractal prediction module stores the prediction model as described in claim 3; The results output module is configured to output the pathological classification prediction results calculated by the prediction model based on the abundance information.

6. The application of reagents for detecting oligosaccharide chain marker combinations in the preparation of kits for assisting in the identification of pathological subtypes of multiple myeloma, characterized in that, The oligosaccharide chain marker combination consists of 10 specific oligosaccharide chains: NGA2F, NG1A2F, NG1A2FB, NA2B, NA2F, NA2FB, NA3, NA3Fb, NA4, and NA4Fb; the pathological subtypes include IgG type, IgA type, light chain type, IgD type, and IgM type.

7. A kit for assisting in the identification of pathological subtypes of multiple myeloma, characterized in that, It contains reagents for detecting combinations of oligosaccharide chain markers; the combinations of oligosaccharide chain markers consist of 10 specific oligosaccharide chains: NGA2F, NG1A2F, NG1A2FB, NA2B, NA2F, NA2FB, NA3, NA3Fb, NA4, and NA4Fb; the pathological subtypes include IgG type, IgA type, light chain type, IgD type, and IgM type.

Images