Application of Raman spectrum as marker in non-invasive identification of multiple myeloma

Abstract

The invention relates to a biomedical detection technology, and discloses application of a Raman spectrum as a marker in non-invasive identification of multiple myeloma. By analyzing Raman spectrum data of a serum sample, specific biomolecular changes related to the MM, including Raman peak positions of nucleic acid, protein, lipids, beta-carotene, saccharides and collagen, are successfully identified. 726, 781, 1579 cm <-1 > corresponding nucleic acids; 621, 643, 759, 1603, 1616 cm <-1 > corresponding proteins; the corresponding lipids are at least one of the first, second, third, fourth, fifth, sixth, seventh, seventh, seventh 957 cm <-1 > corresponds to beta-carotene; 920, 1123 cm <-1 > corresponding saccharides; 1345cm <-1 > corresponds to collagen, and the intensity change of the peak positions above provides a biomarker for the diagnosis of the MM. The biomarker has a diagnostic value through analysis and confirmation. The method not only improves the accuracy of early diagnosis of the MM, but also reduces the detection cost, and has important clinical application value and market potential.

Description

Technical Field

[0001] The present invention relates to the field of medical diagnosis, and in particular to Raman spectroscopy markers for non-invasively identifying multiple myeloma (MM) and applications thereof. Background Art

[0002] Multiple myeloma (MM) is a malignant tumor originating from plasma cells in the bone marrow. Its classification includes IgA, IgG, IgM, IgD, IgE, free light chain (FLC), biclonal and non-secretory types. Traditional detection methods for MM mainly include blood and bone marrow examinations and cytological examinations. The various subtypes of MM are easily confused, so exploring a low-cost diagnostic method that helps to identify MM early is of great significance in the diagnosis and research of hematological diseases.

[0003] In clinical practice, traditional diagnostic methods for MM require high time and economic costs and are invasive, and there is no "gold standard" for diagnosis of some types of the disease. In addition, there is a lack of sufficient evidence for the association between serological indicators related to glucose and lipid metabolism and the typing of MM, which fails to provide sufficient information for the rapid and early identification of MM. Therefore, the development of a rapid identification method for MM based on Raman spectroscopy without antibody labeling and in-depth exploration of the massive serological examination data of MM patients are of great value for the early identification of the disease, which will also improve the diagnostic efficiency of the disease, reduce the cost of testing, and promote the further development of rapid and accurate diagnosis of MM.

[0004] This application uses Raman spectroscopy combined with multivariate analysis to screen the Raman peaks that contribute mainly to disease classification as biomarkers for MM. This marker can distinguish healthy controls, IgA patients, IgG patients, FLC patients, and IgA patients of different subtypes and IgG patients of different subtypes. This application lays the foundation for better use of clinical serological test data for rapid early identification of MM. Summary of the invention

[0005] The first aspect of the present invention provides a method for detecting the Raman peaks 1579, 726, 781 cm representing nucleic acid in peripheral blood based on Raman spectroscopy. -1 , representing protein Raman peaks 621, 643, 759, 1603, 1616 cm -1 , representing the Raman peaks of sugars at 920 and 1123 cm -1 , representing the Raman peak of collagen at 1345 cm -1 and the Raman peaks representing lipids at 1437, 1443, and 1446 cm -1 The use of the characteristic peak intensity level as a biomarker in the preparation of a product for predicting or detecting multiple myeloma.

[0006] Furthermore, the application is to diagnose the progression of the disease and evaluate the condition by detecting the peak intensity of the Raman peak biomarker in the peripheral blood of the subject and comparing it with the average value of the normal control.

[0007] Furthermore, the multiple myeloma is of IgA, IgG or light chain (free light chain, FLC) type.

[0008] Furthermore, the product is a diagnostic kit, a diagnostic reagent, or a diagnostic device.

[0009] The second aspect of the present invention provides a method for detecting the Raman peak 957 cm representing β-carotene in peripheral blood based on Raman spectroscopy. -1 , representing the Raman peak of nucleic acid at 781 cm -1 , representing sugars at 920 and 1123 cm -1 , representing 1345cm of collagen -1 The invention discloses an application of characteristic peak intensity level as a biomarker in the preparation of a product for typing IgA multiple myeloma, wherein the typing is to divide IgA into IgA-κ type and IgA-λ type.

[0010] Furthermore, the application also includes detecting the Raman peaks 621, 643, 759, 1603 cm of the representative protein in peripheral blood based on Raman spectroscopy. -1 , representing the Raman peak of nucleic acid at 1579 cm -1 , and the peaks representing lipids are at 1437, 1443, and 1446 cm -1 The characteristic peak intensity level is used as a biomarker.

[0011] The third aspect of the present invention provides a method for detecting the Raman peak 957 cm representing β-carotene in peripheral blood based on Raman spectroscopy. -1 , representing the Raman peak of nucleic acid at 781 cm -1 The invention discloses an application of characteristic peak intensity level as a biomarker in the preparation of a product for typing IgG type multiple myeloma, wherein the typing is to divide the IgA type into IgG-κ type and IgG-λ type.

[0012] Furthermore, the application also includes detecting the Raman peaks 643, 759, 1603, 1616 cm of the representative protein of peripheral blood based on Raman spectroscopy. -1 , representing the Raman peak of nucleic acid at 1579 cm -1 , and the peaks representing lipids are at 1437, 1443, and 1446 cm -1 The characteristic peak intensity level is used as a biomarker.

[0013] The fourth aspect of the present invention provides a method for detecting the Raman peaks 1443 and 1446 cm representing lipids in peripheral blood based on Raman spectroscopy. -1 , representing sugars at 920 and 1123 cm -1 , representing 1345cm of collagen -1 The invention discloses an application of characteristic peak intensity level as a biomarker in the preparation of a product for typing FLC-type multiple myeloma, wherein the typing is to classify IgA type into FLC-κ type and FLC-λ type.

[0014] Furthermore, the application also includes detecting the Raman peaks 621, 643, 759, 1603, 1616 cm of the representative protein in peripheral blood based on Raman spectroscopy. -1 , representing the Raman peak of nucleic acid at 1579 cm -1 , and the peak representing β-carotene is 957 cm -1 The characteristic peak intensity level is used as a biomarker.

[0015] In the preferred applications of the above four aspects, the detection conditions of Raman spectroscopy are: 785nm laser as excitation light, 40x objective lens, output power of 40mW, ×40 0.75NA Nikon lens is used in the shooting process, the spot diameter range of about 2×2μm on the sample receives the laser beam with an output power of 40mW, the single integration time is 250s, the number of integrations is one, and the measurement range is 600-1800cm -1 , 5-10 sites were measured in each group, with a resolution of 1cm -1 .

[0016] Beneficial effects of the present invention:

[0017] 1. Non-invasive detection: Through Raman spectroscopy technology, the present invention provides a non-invasive detection method, avoiding the discomfort and risks that may be caused by traditional invasive detection methods.

[0018] 2. Rapid diagnosis: The technology of the present invention can quickly analyze serum samples, which helps to accelerate the diagnosis process of MM and is crucial for early detection and treatment.

[0019] 3. High specificity and sensitivity: Raman spectroscopy technology shows high specificity and sensitivity to distinguish MM patients from healthy controls, as well as different subtypes of MM patients, providing the possibility for personalized treatment and prognosis assessment.

[0020] 4. The discovery of specific peaks in Raman spectroscopy provides a basis for future biomarker research and application: The present invention identifies potential biomarkers related to MM, including specific Raman peak positions of nucleic acids, proteins, lipids, β-carotene, carbohydrates and collagen.

[0021] 5. The application of multivariate analysis provides a guarantee for the stability and accuracy of the diagnostic model of this application: through multivariate statistical analysis methods such as OPLS-DA, the study is able to process complex spectral data and improve the accuracy and reliability of data analysis.

[0022] 6. Assistance in clinical diagnosis: The technology of the present invention can assist clinicians in making more accurate judgments when diagnosing MM, especially in the early stages of the disease.

[0023] 7. Reduce testing costs and promote precision medicine: Compared with traditional detection methods, Raman spectroscopy technology may have lower costs, help reduce medical expenses, and help promote the development of precision medicine. Through accurate detection and diagnosis, it can provide patients with more personalized treatment plans.

[0024] In summary, the research results of the present invention provide a new approach for future clinical applications and research. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Schematic diagram of Raman measurement of serum.

[0026] Figure 2 is the average serum spectra of the control group and MM patient group, Figure 2 a From bottom to top, the average serum spectra of the control group, IgA group, IgG group and FLC group; Figure 2 b From bottom to top, the average serum spectra of the control group, IgA-λ group, and IgA-κ group; Figure 2 c From bottom to top, the average serum spectra of the control group, IgG-λ group and IgG-κ group, Figure 2 d is the average serum spectra of the control group, IgA-λ group, and IgA-κ group.

[0027] Figure 3 OPLS-DA identification of control group and MM serum samples, Figure 3 a is the OPLS-DA identification score diagram of the control group, IgA, IgG and FLC drawn with Hotelling's 95% confidence ellipse, Figure 3 b is the differential load line diagram of the control group, IgA, IgG and FLC, Figure 3 c is the V+S diagram of the control group, IgA, IgG and FLC identification, Figure 3 d is the OPLS-DA discrimination score diagram of the control group and IgA group drawn using Hotelling’s 95% confidence ellipse. Figure 3 e is the load line diagram of the control group and IgA identification, f is the V+S diagram of the control group and IgA identification, Figure 3g is the OPLS-DA discrimination score diagram of the control group and IgG drawn with Hotelling’s 95% confidence ellipse, Figure 3 h is the load line diagram of the control group and IgG, Figure 3 i is the V+S diagram for the control group and IgG identification, Figure 3 j is the OPLS-DA discrimination score diagram of the control group and FLC drawn with Hotelling's 95% confidence ellipse, Figure 3 k is the line graph of the control group and FLC discrimination loadings drawn using Hotelling's 95% confidence ellipse, Figure 3 L is the V+S diagram of the control group and FLC.

[0028] Figure 4 OPLS-DA identification of serum samples from the control group and IgA group. Figure 4 A is the OPLS-DA identification score diagram of the control group, IgA-λ, and IgA-κ drawn using Hotelling's 95% confidence ellipse. Figure 4 B is the differential load line diagram of the control group, IgA-λ, and IgA-κ. Figure 4 C is the V+S diagram of the control group, IgA-λ, and IgA-κ. Figure 4 D is the OPLS-DA discrimination score diagram of the control group and IgA-λ drawn with Hotelling's 95% confidence ellipse. Figure 4 E is the load line diagram of the control group and IgA-λ, Figure 4 F is the V+S diagram of the control group and IgA-λ identification, Figure 4 G is the OPLS-DA discrimination score diagram of the control group and IgA-κ drawn with Hotelling's 95% confidence ellipse, Figure 4 H is the load line diagram of the control group and IgA-κ, Figure 4 I is the V+S diagram of the control group and IgA-κ identification, Figure 4 J is the identification score diagram of IgA-λ and IgA-κ by OPLS-DA drawn with Hotelling's 95% confidence ellipse, Figure 4 K is the IgA-λ and IgA-κ discrimination load line diagram, Figure 4 L is the V+S diagram for distinguishing IgA-λ and IgA-κ.

[0029] Figure 5 OPLS-DA identification of serum samples from the control group and IgG group. Figure 5 A is the OPLS-DA identification score diagram of the control group, IgG-λ, and IgG-κ drawn using Hotelling's 95% confidence ellipse. Figure 5 B is the load line diagram of the control group, IgG-λ, and IgG-κ. Figure 5C is the V+S diagram for the control group, IgG-λ, and IgG-κ identification. Figure 5 D is the OPLS-DA discrimination score diagram of the control group and IgG-λ drawn with Hotelling's 95% confidence ellipse. Figure 5 E is the load line diagram for the control group and IgG-λ, Figure 5 F is the V+S diagram for the control group and IgG-λ identification, Figure 5 G is the OPLS-DA discrimination score diagram of the control group and IgG-κ drawn with Hotelling's 95% confidence ellipse, Figure 5 H is the load line diagram for the control group and IgG-κ, Figure 5 I is the V+S diagram for the control group and IgG-κ identification, Figure 5 J is the OPLS-DA discrimination score diagram of IgG-λ and IgG-κ drawn with Hotelling’s 95% confidence ellipse, Figure 5 K is the loading line diagram for the discrimination between IgG-λ and IgG-κ, and l is the V+S diagram for the discrimination between IgG-λ and IgG-κ.

[0030] Figure 6 OPLS-DA identification of serum samples from the control group and FLC group. Figure 6 A is the OPLS-DA identification score diagram of the control group, FLC-λ, and FLC-κ drawn using Hotelling's 95% confidence ellipse. Figure 6 B is the differential loading line diagram of the control group, FLC-λ, and FLC-κ. Figure 6 C is the V+S diagram for the identification of the control group, FLC-λ, and FLC-κ. Figure 6 D is the OPLS-DA discrimination score diagram of the control group and FLC-λ drawn with Hotelling's 95% confidence ellipse. Figure 6 E is the load line diagram for the control group and FLC-λ, Figure 6 F is the V+S diagram of the control group and FLC-λ, Figure 6 G is the OPLS-DA discrimination score diagram of the control group and FLC-κ drawn with Hotelling's 95% confidence ellipse, Figure 6 H is the differential loading line diagram of the control group and FLC-κ, Figure 6 I is the V+S diagram for identification of the control group and FLC-κ, Figure 6 J is the identification score diagram of FLC-λ and FLC-κ by OPLS-DA drawn with Hotelling's 95% confidence ellipse, Figure 6 K is the FLC-λ and FLC-κ discrimination load line diagram, Figure 6 L is the V+S diagram for distinguishing FLC-λ and FLC-κ.

[0031] Figure 7Potential biomarker statistics and serum biochemical results for control group vs IgA vs IgG vs FLC, Figure 7 A is the control group vs IgA vs IgG vs FLC potential biomarkers, Figure 7 B is total protein, Figure 7 C is glucose, Figure 7 D stands for high-density lipoprotein.

[0032] Figure 8 The potential biomarker statistics and serum biochemical results of the control group vs IgA group are shown in Figure 2. Figure 8 A is the potential biomarker of the control group vs IgA group, Figure 8 B is total protein, Figure 8 C is glucose, Figure 8 D stands for high-density lipoprotein.

[0033] Fig. 9 Statistics of potential biomarkers and serum biochemical results of the control group vs IgG group. Fig. 9 A is the potential biomarker of the control group vs IgG group, Fig. 9 B is the total protein graph, 9C is the glucose graph, and 9D is the high-density lipoprotein.

[0034] Fig.10 The potential biomarker statistics and serum biochemical results of the control group vs FLC group are shown in Figure 2. Fig.10 A is the potential biomarker of the control group vs IgG group, Fig.10 B is the total protein graph, 10C is the glucose graph, and 10D is the high-density lipoprotein.

[0035] Fig.11 To establish the differentiation model between the control group and the IgA, IgG, and FLC groups, the control group and the IgA group, the control group and the IgG group, and the control group and the FLC group, Fig.11 A is the arrangement diagram of the control group, IgA group, IgG group and FLC group displayed by OPLS-DA. Fig.11 B is the cluster analysis of four groups. Fig.11 C is the ROC curve of the four groups. Fig.11 D is the arrangement diagram of the control group and IgA group. Fig.11 E is the cluster analysis of the control group and IgA group. Fig.11 F is the ROC diagram of the control group and IgA group. Fig.11 G is the arrangement diagram of the control group and IgG group, Fig.11 H is the cluster analysis of the control group and IgG group, Fig.11 I is the ROC diagram of the control group and IgG group, Fig.11 J is the arrangement diagram of the control group and the FLC group, Fig.11K is the cluster analysis diagram of the control group and FLC group, Fig.11 L is the ROC diagram of the control group and the FLC group.

[0036] Fig.12 To establish the differentiation model between the control group and the IgA-λ, IgA-κ group, the control group and the IgA-λ group, the control group and the IgA-κ group, and the IgA-λ and IgA-κ group, Fig.12 A is the arrangement diagram of the control group, IgA-λ group and IgA-κ group displayed by OPLS-DA. Fig.12 B is the cluster analysis of three groups. Fig.12 C is the ROC curve of the three groups. Fig.12 D is the arrangement diagram of the control group and IgA-λ group, Fig.12 E is the cluster analysis of the control group and IgA-λ group. Fig.12 F is the ROC diagram of the control group and IgA-λ group, Fig.12 G is the arrangement diagram of the control group and IgA-κ group, Fig.12 H is the cluster analysis of the control group and IgA-κ group. Fig.12 I is the ROC diagram of the control group and IgA-κ group, Fig.12 J is the arrangement diagram of IgA-λ group and IgA-κ group, Fig.12 K is the cluster analysis diagram of IgA-λ group and IgA-κ group, Fig.12 L is the ROC diagram of the IgA-λ group and the IgA-κ group.

[0037] Fig.13 To establish a distinguishing model between the control group and the IgG-λ, IgG-κ group, the control group and the IgG-λ group, the control group and the IgG-κ group, and the IgG-λ and IgG-κ group, Fig.13 A is the arrangement diagram of the control group, IgG-λ group and IgG-κ group displayed by OPLS-DA. Fig.13 B is the cluster analysis of three groups. Fig.13 C is the ROC curve of the three groups. Fig.13 D is the arrangement diagram of the control group and IgG-λ group, Fig.13 E is the cluster analysis of the control group and IgG-λ group. Fig.13 F is the ROC diagram of the control group and IgG-λ group. Fig.13 G is the arrangement diagram of the control group and IgG-κ group, Fig.13 H is the cluster analysis of the control group and IgG-κ group. Fig.13 I is the ROC diagram of the control group and IgG-κ group, Fig.13 K is the cluster analysis diagram of IgG-λ group and IgG-κ group, Fig.13 L is the ROC plot of the IgG-λ group and the IgG-κ group.

[0038] Fig.14 To establish the distinguishing model between the control group and the FLC-λ group, the FLC-κ group, the control group and the FLC-λ group, the control group and the FLC-κ group, and the FLC-λ and FLC-κ group, Fig.14 A is the arrangement diagram of the control group, FLC-λ group and FLC-κ group displayed by OPLS-DA, Fig.14 B is the cluster analysis of three groups. Fig.14 C is the ROC curve of the three groups. Fig.14 D is the arrangement diagram of the control group and FLC-λ group, Fig.14 E is the cluster analysis of the control group and the λ group, Fig.14 F is the ROC diagram of the control group and FLC-λ group, Fig.14 G is the arrangement diagram of the control group and FLC-κ group, Fig.14 H is the cluster analysis of the control group and FLC-κ group, Fig.14 I is the ROC diagram of the control group and FLC-κ group, Fig.14 J is the arrangement of FLC-λ group and FLC-κ group, Fig.14 Cluster analysis diagram of FLC-λ group and FLC-κ group, Fig.14 L is the ROC diagram of the FLC-λ group and the FLC-κ group. Example:

[0039] Unless otherwise defined, the technical terms used in the following examples have the same meanings as those generally understood by those skilled in the art to which the present invention belongs. The test reagents used in the following examples are conventional biochemical reagents unless otherwise specified; the experimental methods are conventional methods unless otherwise specified. The present invention is described in detail below in conjunction with the embodiments and the accompanying drawings.

[0040] Example 1 Serum and Raman spectroscopy data collection

[0041] 1. Experimental Group Establishment

[0042] 35 patients who visited the Hematology Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences) in 2022 were recruited, including 19 males and 16 females, with a youngest age of 38 and a oldest age of 76. They were divided into three groups of MM patients and one control group: 10 patients in the IgA group, including 5 IgA-λ patients and 5 IgA-κ patients; 14 patients in the IgG group, including 5 IgG-λ patients and 9 IgG-κ patients; 11 patients in the FLC group, including 6 λ patients and 5 κ patients; 13 healthy people served as the control group.

[0043] The above patient cases were diagnosed by experienced hematologists after blood phase, bone marrow phase, cytogenetics, immunological phenotype analysis and gene analysis. This study was approved by the Ethics Committee of the Blood Disease Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences) (KT2020016-EC-2). The serum samples used in the study were the remaining samples of clinical tests in MM patients and the control group. All MM patients and the control group underwent routine serum biochemical tests. The serum biochemical test data came from the Clinical Testing Center of the Blood Disease Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, Chinese Academy of Medical Sciences).

[0044] 2. Serological testing

[0045] The general clinical data of the subjects were collected by conventional serological testing methods. Serum was collected from the subjects after fasting for 10 hours, and the peripheral blood total protein (TP), glucose, triglyceride (TG), total cholesterol (TC), high-density lipoprotein (HDL) and low-density lipoprotein (LDL) were measured by automatic biochemical analyzer.

[0046] 3. Peripheral serum Raman spectroscopy analysis

[0047] Raman spectroscopy was used to analyze the peripheral blood of the subjects. The specific method was as follows: 5 μL of serum was dropped on a quartz glass slide, and the confocal Raman spectrometer XploRA Raman microscope was used to measure the amount. A 785 nm laser was selected as the excitation light, the output power was 40 mW, the objective lens was selected at 40 times, and the specimen was fixed on an XYZ three-dimensional platform. A ×40 0.75NA Nikon lens was used in the shooting process, and a spot size range of about 2×2 μm on the sample was irradiated by a laser beam with an output power of 40 mW. The single integration time was 250 s, the number of integrations was once, the measurement range was 600-1800 cm-1, 5-10 sites were measured in each group, and the resolution was 1 cm -1 At the same time, the Raman spectrum of the quartz glass slide was measured as the background. Labspec6 software was used for data processing such as smoothing, background removal and baseline correction. All spectra were analyzed at their respective 1450cm -1 The Raman peaks were intensity normalized to the internal standard.

[0048] 4. Experimental results:

[0049] Serological results:

[0050] Table 1. Comparison of clinical data between IgA, IgG, FLC and control groups

[0051]

[0052]

[0053] Raman spectroscopy results:

[0054] In order to study the Raman spectral characteristics of serum of MM patients and control group, this embodiment obtained 73, 66, 76 and 57 Raman characteristic spectra of serum of control group, IgA group, IgG group and FLC group respectively, among which the IgA group included 29 spectra of IgA-λ patients and 37 spectra of IgA-κ patients. The IgG group included 27 spectra of IgG-λ patients and 49 spectra of IgG-κ patients. The FLC group included 33 spectra of IgA-λ patients and 24 spectra of κ patients.

[0055] Depend on Figure 2 A-2D shows that in the range of 600-1800cm -1 Serum Raman spectra of the control group, IgA, IgG and FLC group, control group and IgA group, control group and IgG group, control group and FLC group are within the range of . The peak positions of the relevant serum Raman spectra are assigned in Table 2.

[0056] Table 2 Peak positions of serum Raman spectra

[0057]

[0058]

[0059] Figure 2 A shows the Raman spectra of the control group, IgA, IgG and FLC samples. The red, blue, green, purple, orange and black vertical lines in the figure represent the Raman spectra of nucleic acids (726, 781, 1579 cm -1 ), protein (621, 643, 759, 1603, 1616cm -1 ), β-carotene (957cm- 1 ), lipids (1437, 1443, 1446cm -1 ), sugars (920, 1123cm -1 ) and collagen (1345cm -1 ) related peak positions.

[0060] Figure 2 B shows the Raman spectra of the control group, IgA-λ, and IgA-κ. Figure 2 C shows the Raman spectra of the control group, IgG-λ and IgG-κ. Figure 2 D shows the Raman spectra of the control group, IgG-λ, and IgG-κ, and the spectral graphs show similar morphology.

[0061] It can be seen from the above figures that it is difficult to identify the differences in serum substances between the MM patient group and the control group based on the above spectral patterns and peak positions.

[0062] Example 2 Establishment of OPLS-DA diagnostic model based on Raman spectroscopy data analysis

[0063] The present invention demonstrates in Example 2 a classification model established in combination with the OPLS-DA method, and further screens out peak positions that can effectively distinguish between the control group and the MM patient group as a means of potential biomarkers.

[0064] 1. Experimental Methods

[0065] 1.1 Method of establishing OPLS-DA diagnostic model

[0066] The present invention forms a complete set of MM identification research ideas from serum sample collection, Raman spectroscopy detection, multivariate analysis to the establishment of identification models (see Figure 1 ).

[0067] OPLS-DA (orthogonal partial least-squares discrimination analysis) is a multivariate analysis method that can be used in the analysis of multidimensional data such as spectroscopy and metabolomics. OPLS analysis is based on the following principles: OPLS removes systematic orthogonal variables from a given data set X and distinguishes these orthogonal variables from non-orthogonal variables, which can be analyzed separately. The OPLS method uses the information in the response variable Y to divide X into three parts. That is,

[0068] X=TPPTP+TOPTO+E

[0069] Among them, TP represents the predicted score matrix of X, PTP represents the predicted load matrix of X, TPPTP represents the predicted part, TO represents the score matrix of the orthogonal components of X and Y (called OPLS components), PTO represents the corresponding load matrix, TOPTO represents the part orthogonal to Y, and E is the residual matrix.

[0070] The implementation of the OPLS method is completed in two steps:

[0071] In the first step, the variables orthogonal to Y are removed from the X data matrix, that is,

[0072] XP=X-TOPTO

[0073] Where TO is the score matrix of the orthogonal components of Y, and PTO is the corresponding loading matrix.

[0074] The second step is to perform partial less squares analysis on XP.

[0075] The present invention uses SIMCA14.1 software to perform supervised orthogonal partial least squares discrimination analysis (OPLS-DA) on the serum Raman spectroscopy data of MM patients and the control group. The performance of the OPLS model is evaluated using the goodness of fit parameters R2 and Q2. Under the null hypothesis, the model is resampled 200 times by random changes in the y matrix to perform model validation. In order to find the Raman peak positions with statistically significant differences in the classification model as potential biomarkers, we used cluster analysis and V+S analysis. Based on comprehensive consideration of parameters such as correlation coefficient, load, and distance from the center in the V+S diagram, the peak position with VariableImportance (VIP)>0.5 is selected as a potential biomarker. The potential biomarkers obtained are subjected to a significance test, and the potential biomarkers with P<0.05 are considered to be statistically significant. Origin software is used for relevant data processing. IBM SPSS Statistics 27.0 and Graphpad Prism 9.0 are used for statistical analysis, and statistical correlation graphs are drawn.

[0076] 1.2 Verification of the effectiveness of the OPLS-DA model

[0077] The effectiveness of the supervised OPLS-DA model based on multi-parameter analysis can be demonstrated by permutation plots ( Fig.11 A, 11D, 11G, 11J), cluster analysis diagram ( Fig.11 B, 11E, 11H, 11K) and ROC plots( Fig.11 C, 11F, 11I, 11L) for evaluation.

[0078] 1.2.1Permutation plot: In order to determine whether the OPLS-DA model is valid, permutation is used to determine whether the model is valid. Permutation analysis shows that the intercept of Q2 on the Y axis is negative, indicating that the OPLS-DA model is valid and not overfitting.

[0079] 1.2.2 Cluster analysis: Cluster analysis is an unsupervised learning method. Its principle is to divide the samples in the data set into different groups with similar characteristics, called "clusters", according to the proximity or similarity between the samples. The similarity within the cluster is high, while the similarity between different clusters is low. 1.2.3 ROC plots: ROC plots are used to evaluate the authenticity of the identification method. The closer the Area Under Curve (AUC) is to 1, the higher the authenticity of the identification method. The receiver operating characteristic (ROC) curve is a comprehensive indicator reflecting the continuous variables of sensitivity and specificity. It uses the composition method to reveal the relationship between sensitivity and specificity. It calculates a series of sensitivities and specificities by setting multiple different critical values ​​for the continuous variables, and then draws a curve with sensitivity as the ordinate and (1-specificity) as the abscissa. The larger the area under the curve (AUC), the higher the diagnostic accuracy. On the ROC curve, the point closest to the upper left of the coordinate graph is the critical value with high sensitivity and specificity.

[0080] True positive rate (TPR) and false positive rate (FPR) are two indicators used to evaluate the ROC curve. Add the horizontal and vertical coordinates of (b) and (e) and re-create the combined graph.

[0081] True positive rate: false positive rate: Among them, True Positive (True Positive, TP): predict the positive class as the positive class number. True Negative (True Negative, TN): predict the negative class as the negative class number. False Positive (False Positive, FP): predict the negative class as the positive class number → false positive (Type I error). False Negative (False Negative, FN): predict the positive class as the negative class number → missed report (Type II error). Intuitively, TPR represents the probability of correctly classifying a positive example, and FPR represents the probability of misclassifying a negative example as a positive example. In the ROC space, the horizontal axis of each point is FPR and the vertical axis is TPR, which depicts the trade-off2 of the classifier between TP (True Positive Rate) and FP (False Positive Rate).

[0084] 2 Experimental results:

[0085] The OPLS-DA model and multi-parameter analysis were used to initially screen potential biomarkers, as follows:

[0086] 2.1 Potential Raman spectroscopy biomarkers in the initial screening of control group vs IgA group vs IgG group vs FLC group

[0087] In order to establish a method for identifying the control group, IgA group, IgG group and FLC group based on multi-parameter analysis, the Raman spectral data of the control group, IgA group, IgG group and FLC group were respectively used to form four groups of data. SIMCA-P software was used to apply supervised orthogonal partial least squares discriminant analysis (OPLS-DA) to the sample data and a detailed analysis and comparison was carried out.

[0088] The effectiveness of the established supervised OPLS-DA model was verified by permutation plots, cluster analysis plots, and ROC plots. The permutation plots results are shown in ( Fig.11 A, 11D, 11G, 11J).

[0089] The cluster analysis results are shown in ( Fig.11 B, 11E, 11H, 11K). Permutation analysis shows that the intercept of Q2 on the Y axis is negative, indicating that the OPLS-DA model is valid and not overfitting ( Fig.11 A, 11D, 11G, 11J). The cluster analysis results show that the model has a good distinguishing effect on each group of samples.

[0090] ROC curve shows ( Fig.11 C, 11F, 11I, 11L), from the ROC diagrams of these different group comparisons, it can be seen that in the control group vs IgA group vs IgG group vs FLC group model, AUC (control group) = 1, AUC (IgA group) = 1, AUC (IgG group) = 1, and AUC (FLC group) = 1. In the control group vs IgA group model, AUC (control) = 1, AUC (IgA) = 1 ( Fig.11 F), in the control group vs Ig group G model, AUC(control)=1, AUC(IgG)=1( Fig.11 I), in the control group vs FLC model, AUC(control)=1, AUC(FLC)=1( Fig.11 L), indicating that the discriminant analysis results are highly accurate.

[0091] In order to further study the relationship between Raman spectral data of the control group, IgA group, IgG group and FLC group and MM and screen potential biomarkers, the OPLS-DA model was used to further analyze the relationship between different groups. Figure 3A is the OPLS-DA scoreplot. In the figure, the four groups of samples are clearly divided into groups. The control group and the FLC group are located on the positive half axis of X, and the IgA group and the IgG group are located on the negative half axis of X, reflecting that the control group, IgA group, IgG group and FLC group are distinguished. This distinction result shows that OPLS-DA can well identify the serum spectral data of patients in the control group, IgA group, IgG group and FLC group, laying the foundation for analyzing the material characteristics of the four groups.

[0092] Figure 3 B is the OPLS-DA loading plot, which is used to preliminarily screen the Raman peaks that contribute to the identification model of the control group vs. IgA group vs. IgG group vs. FLC group. The peak numbers in red, blue, green, purple, orange, and black in the figure are related to nucleic acids, proteins, β-carotene, lipids, carbohydrates, and collagen, respectively. The loading plot and the score plot are correlated in representing the relationship between the sample substance content, that is, the substance represented by the peak located on the positive semi-axis of the vertical coordinate of the loading plot has a relatively higher content in the group on the positive semi-axis of the horizontal coordinate of the score plot, and the negative semi-axis of the vertical coordinate of the loading plot and the negative semi-axis of the horizontal coordinate of the score plot also have a similar corresponding relationship. The characteristic peaks representing nucleic acid (726, 1579 cm-1), protein (621, 643, 759, 1603, 1616 cm-1), lipid (1437, 1443, 1446 cm-1), β-carotene (957 cm-1), carbohydrate (920, 1123 cm-1) and collagen (1345 cm-1) played an important role in the identification of the four groups of samples, reflecting that the contents of nucleic acid, protein, lipid and β-carotene in the control group were higher than those in the MM group ( Figure 3 B). Figure 3 C is OPLS-DA V+S plot, which integrates VIP and correlation coefficient, which characterize the contribution of peak positions to the classification model, and is used to further screen potential biomarkers. In subsequent analysis, the characteristic peak positions are tested for significance, and the peak positions that can effectively identify the control group and MM are determined as potential biomarkers. V+S plot can provide a list of Raman peak positions in descending order of VIP values, and peak positions with VIP>0.5 and biological significance are screened as potential biomarkers.

[0093] 2.2 Potential Raman spectroscopy biomarkers in the initial screening of control group vs IgA group, control group vs IgG group, and control group vs FLC group Based on the control group vs IgA vs IgG vs FLC model, the control group was combined with the samples of the IgA group, IgG group, and FLC group for OPLS-DA analysis ( Figure 3 DL). Figure 3 D, 3G, and 3J are the OPLS-DA score plots of the three models: control group vs IgA group, control group vs IgG group, and control group vs FLC group. The two groups of samples in the three figures are located on the positive and negative half axes of the X-axis, respectively. The samples are clearly clustered in the scatter plots, reflecting that OPLS-DA can better extract the difference information in the spectrum. The established identification method can identify the differences in the metabolic components of serum samples. The three models can well identify the two groups of samples in the model ( Figure 3 D, 3G, 3J).

[0094] Figure 3 E, 3H, and 3K are the loading plots of the three models of control vs IgA, control vs IgG, and control vs FLC, respectively. The figures show that the peak intensities representing proteins, nucleic acids, and lipids in the control group are generally higher than those in the MM group, while the peak intensities representing carbohydrates and collagen are lower than those in the MM group ( Figure 3 Specifically, compared with the control group, the intensities of the Raman peaks representing nucleic acids (726, 781, 1579 cm-1), proteins (621, 643, 759, 1603, 1616 cm-1), lipids (1437, 1443, 1446 cm-1), and β-carotene (957 cm-1) in MM patients were significantly reduced, and the intensities of the Raman peaks representing carbohydrates (920, 1123 cm-1) and collagen (1345 cm-1) were significantly increased. Figure 3 F, 3I, and 3L are the V+S plots of the control group vs IgA group, control group vs IgG group, and control group vs FLC group, respectively. The V+S plot can provide a list of Raman peaks in descending order of VIP values. This list provides the main basis for determining potential biomarkers in the control, IgA, IgG, and FLC models ( Figure 3 C, 3F, 3I, 3L). Peaks with VIP>0.5 and biological significance were screened as potential biomarkers. In the four models of control vs IgA vs IgG vs FLC, control vs IgA, control vs IgG, and control vs FLC, VIP (VIP>0.5), correlation coefficient, load, distance from the center in the V+S graph and other related parameters were comprehensively considered, and important Raman peaks affecting sample classification were found. In the subsequent biomarker verification process, Raman peaks with no statistical difference were excluded from the biomarker range.

[0095] 2.2.1 Potential Raman spectroscopy biomarkers in control vs IgA group After screening out potential IgA biomarkers, the IgA group model was used to analyze the differences in sample substance content and screen for potential biomarkers in the IgA group. The effectiveness of the supervised OPLS-DA model established based on multi-parameter analysis can be demonstrated by permutation plots ( Fig.12 A, 12D, 12G, 12J), cluster analysis diagram ( Fig.12 B,12E,12H,12K) and ROC plots( Fig.12 C, 12F, 12I, 12L). Permutation analysis showed that the intercept of Q2 on the Y axis was negative, indicating that the OPLS-DA model was established and not overfitting ( Fig.12 A, 12D, 12G, 12J). The cluster analysis results showed that the model had a good discriminative effect on each group of samples. The ROC curve showed that in the control vs IgA-λ vs IgA-κ model, AUC(control) = 1, AUC(IgA-λ) = 1, AUC(IgA-κ) = 1 ( Fig.12 C), in the control vs IgA-λ model, AUC(control)=1, AUC(IgA-λ)=1( Fig.12 F), in the control vs IgA-κ model, AUC(control)=1, AUC(IgA-κ)=1( Fig.12 I), in the IgA-λ vs IgA-κ model, AUC(IgA-λ)=1, AUC(IgA-κ)=1( Fig.12 L), indicating that the discriminant analysis results are highly accurate.

[0096] Figure 4 A is the OPLS-DA score plot. The three groups of samples are clearly divided into groups. The control group is located on the positive half axis of Y, and the IgA-λ and IgA-κ groups are located on the negative half axis of Y, reflecting that the control group and the IgA group are distinguished. This distinction result shows that OPLS-DA can well distinguish the serum spectral data of the control, IgA-λ and IgA-κ patients, providing conditions for analyzing the material characteristics of the three groups. Figure 4 B is the OPLS-DA loading plot, which is used to preliminarily screen the Raman peaks that contribute to the identification model of the control vs IgA group. The red, blue, green, purple, orange and black peak numbers in the figure are related to nucleic acids, proteins, β-carotene, lipids, carbohydrates and collagen, respectively. -1 ), protein (621, 643, 759, 1603, 1616cm -1 ), lipids (1437, 1443, 1446cm -1), β-carotene (957cm-1), sugars (920, 1123cm -1 ) and collagen (1345cm -1 ) played an important role in the identification of the four groups of samples, reflecting that the nucleic acid, protein, lipid and β-carotene contents of the control group were higher than those of the IgA group ( Figure 4 B). Figure 4 C is the OPLS-DA V+S plot, which was used to further screen potential biomarkers. In the subsequent analysis, the characteristic peak positions were tested for significance, and the peak positions that could effectively distinguish the control group and the IgA group were determined as potential biomarkers.

[0097] Based on the control vs IgA group model, the control and two IgA group samples were combined in pairs for OPLS-DA analysis ( Figure 4 DL). Figure 4 D, 4G, and 4J are the OPLS-DA score plots of the three models: control vs IgA-λ, control vs IgA-κ, and IgA-λ vs IgA-κ. The two groups of samples in the three figures are located on the positive and negative half axes of the X-axis, respectively. The samples are clearly clustered in the scatter plots, reflecting that the three models have good discrimination capabilities for the two groups of samples in the models. Figure 4 E, 4H, and 4K are the loading plots of the control vs IgA-λ, control vs IgA-κ, and IgA-λ vs IgA-κ models, respectively. The figures show that the peak intensities representing proteins, nucleic acids, and lipids in the control group are generally higher than those in the IgA group, while the peak intensities representing carbohydrates and collagen are lower than those in the IgA group ( Figure 3 Specifically, compared with the control group, the intensities of the Raman peaks representing nucleic acids (726, 781, 1579 cm-1), proteins (621, 643, 759, 1603, 1616 cm-1), lipids (1437, 1443, 1446 cm-1), and β-carotene (957 cm-1) in the IgA group were significantly reduced, and the intensities of the Raman peaks representing carbohydrates (920, 1123 cm-1) and collagen (1345 cm-1) were significantly increased. Figure 4 F, 4I, and 4L are the V+Splot of the control vs IgA-λ, control vs IgA-κ, and IgA-λ vs IgA-κ models, respectively. The V+Splot provides the main basis for determining the potential biomarkers in the control and IgA group models ( Figure 4C, 4F, 4I, 4L). Based on the V+S plot, a list of peaks with VIP>0.5 can be derived, and peaks with biological significance can be screened out as the peak range of potential biomarkers according to previous literature. In the control vs IgA group, control vs IgA-λ, control vs IgA-κ, and IgA-λvsIgA-κ models, the same method was used to screen and verify biomarkers.

[0098] 2.2.2 Potential Raman spectroscopy biomarkers in the initial screening of control vs IgG groups Based on the screening of IgG potential biomarkers, the IgG group model was used to analyze the differences in sample substance content and screen the potential biomarkers of the IgG group. The effectiveness of the supervised OPLS-DA model established based on multi-parameter analysis can be verified by permutation plots ( Fig.13 A, 13D, 13G, 13J), cluster analysis diagram ( Fig.13 B, 13E, 13H, 13K) and ROCplots( Fig.13 C, 13F, 13I, 13L). Permutation analysis showed that the intercept of Q2 on the Y axis was negative, indicating that the OPLS-DA model was established and not overfitting ( Fig.13 A, 13D, 13G, 13J). The cluster analysis results showed that the model had a good discriminative effect on each group of samples. The ROC curve showed that in the control vs IgG-λ vs IgG-κ model, AUC(control) = 1, AUC(IgG-λ) = 1, AUC(IgG-κ) = 1 ( Fig.13 C), in the control vs IgG-λ model, AUC(control)=1, AUC(IgG-λ)=1( Fig.13 F), in the control vs IgG-κ model, AUC(control)=1, AUC(IgG-κ)=1( Fig.13 I), in the IgG-λvsIgG-κ model, AUC(IgG-λ)=1,AUC(IgG-κ)=1( Fig.13 L), indicating that the discriminant analysis results are highly accurate.

[0099] Figure 5 A is the OPLS-DA score plot. In the figure, the three groups of samples are clearly divided into groups. The control group is located on the positive half axis of Y, and the IgG-λ and IgG-κ groups are located on the negative half axis of Y, reflecting that the control group and the IgG group are distinguished. This distinction result shows that OPLS-DA can well distinguish the serum spectral data of the control, IgG-λ and IgG-κ patients, providing conditions for analyzing the material characteristics of the three groups. Figure 5B is the OPLS-DA loading plot, which is used to preliminarily screen the Raman peaks that contribute to the identification model of the control vs IgG group. The red, blue, green, purple, orange and black peak numbers in the figure are related to nucleic acids, proteins, β-carotene, lipids, carbohydrates and collagen, respectively. Figure 5 C is the OPLS-DA V+S plot, which was used to further screen potential biomarkers. In the subsequent analysis, the characteristic peak positions were tested for significance, and the peak positions that could effectively identify the control group and the IgG group were determined as potential biomarkers.

[0100] Based on the control vs IgG group model, the control and two IgG group samples were combined in pairs for OPLS-DA analysis ( Figure 5 DL). Figure 5 D, 5G, and 5J are the OPLS-DA score plots of the three models: control vs IgG-λ, control vs IgG-κ, and IgG-λ vs IgG-κ. The two groups of samples in the three figures are located on the positive and negative half axes of the X-axis, respectively. The samples are clearly clustered in the scatter plots, reflecting that the three models have good discrimination capabilities for the two groups of samples in the models. Figure 5 E, 5H, and 5K are the loading plots of the control vs IgG-λ, control vs IgG-κ, and IgG-λ vs IgG-κ models, respectively. The figures show that the peak intensities representing proteins, nucleic acids, and lipids in the control group are generally higher than those in the IgG group, while the peak intensities representing carbohydrates and collagen are lower than those in the IgG group. Figure 5 F, 5I, and 5L are the V+S plots of the control vs IgG-λ, control vs IgG-κ, and IgG-λ vs IgG-κ models, respectively. The V+S plot provides the main basis for determining the potential biomarkers in the control and IgG group models ( Figure 5 C, 5F, 5I, 5L). Based on the V+S plot, a list of peaks with VIP>0.5 can be derived, and peaks with biological significance can be screened out as the peak range of potential biomarkers based on previous literature. In the control vs IgG group, control vs IgG-λ, control vs IgG-κ, and IgG-λ vs IgG-κ models, the same method was used to screen and verify biomarkers.

[0101] 2.2.3 Potential Raman spectroscopy biomarkers in control vs FLC group Based on the screening of potential biomarkers for FLC, the FLC group model was used to analyze the differences in sample substance content and screen potential biomarkers for the FLC group. The effectiveness of the supervised OPLS-DA model established based on multi-parameter analysis can be demonstrated by permutation plots ( Fig.14A, 14D, 14G, 14J), cluster analysis diagram ( Fig.14 B, 14E, 14H, 14K) and ROC plots( Fig.14 C, 14F, 14I, 14L). Permutation analysis showed that the intercept of Q2 on the Y axis was negative, indicating that the OPLS-DA model was established and not overfitting ( Fig.14 A, 14D, 14G, 14J). The cluster analysis results showed that the model had a good discriminative effect on each group of samples. The ROC curve showed that in the control vsλvsκ model, AUC(control)=1, AUC(λ)=1, AUC(κ)=1( Fig.14 C), in the control vs λ model, AUC(control) = 1, AUC(λ) = 1 ( Fig.14 F), in the control vs. κ model, AUC(control) = 1, AUC(κ) = 1 ( Fig.14 I), in the λvsκ model, AUC(λ)=1,AUC(κ)=1( Fig.14 L), indicating that the discriminant analysis results are highly accurate.

[0102] Figure 6 A is the OPLS-DA score plot. In the figure, the three groups of samples are clearly divided into groups. The control group is located on the positive half axis of X, and the λ and κ groups are located on the negative half axis of X, reflecting that the control group and the FLC group are distinguished. This distinction result shows that OPLS-DA can well distinguish the serum spectral data of the control, λ and κ patients, providing conditions for analyzing the material characteristics of the three groups. Figure 6 B is the OPLS-DA loading plot, which is used to preliminarily screen the Raman peaks that contribute to the identification model of the control vs IgA group. The red, blue, green, purple, orange and black peak numbers in the figure are related to nucleic acids, proteins, β-carotene, lipids, carbohydrates and collagen, respectively. Figure 6 C is the OPLS-DA V+S plot, which was used to further screen potential biomarkers. In the subsequent analysis, the characteristic peak positions were tested for significance, and the peak positions that could effectively distinguish the control group and the FLC group were determined as potential biomarkers.

[0103] Based on the control vs FLC group model, the control and two FLC group samples were combined in pairs for OPLS-DA analysis ( Figure 6 DL). Figure 6 D, 6G, and 6J are the OPLS-DA score plots of the three models: control vs λ, control vs κ, and λ vs κ. The two groups of samples in the three figures are located on the positive and negative half axes of the X-axis, respectively. The samples are clearly clustered in the scatter plots, reflecting that the three models have good discrimination capabilities for the two groups of samples in the model. Figure 6 E, 6H, and 6K are the loading plots of the control vsλ, control vsκ, and λ vsκ models, respectively. The figures show that the peak intensities representing proteins, nucleic acids, and lipids in the control group are generally higher than those in the FLC group, while the peak intensities representing carbohydrates and collagen are lower than those in the FLC group. Figure 6 F, 6I, and 6L are the V+S plots of the control vsλ, control vsκ, and λ vsκ models, respectively. The V+S plot provides the main basis for determining the potential biomarkers in the control and FLC group models ( Figure 6 C, 6F, 6I, 6L). Based on the V+S plot, a list of peaks with VIP>0.5 can be derived, and peaks with biological significance can be screened out according to previous literature as the peak range of potential biomarkers. In the control vs FLC group, control vsλ, control vsκ, and λvsκ models, the same method was used to screen and verify biomarkers.

[0104] Based on the above experiments, the potential Raman spectroscopy biomarkers screened in each group are shown in the following table:

[0105]

[0106] Example 3 Validation of the validity and accuracy of biomarkers using statistical analysis

[0107] Experimental methods: Statistical analysis was used to validate the results of biomarkers. IBM SPSS Statistics 27.0 and Graphpad Prism 9.0, as mentioned above, were used for data analysis and graphics generation. The counted spectral data were independent of each other. Normally distributed data were expressed as M ± SE [mean ± standard error]. Groups with homogeneous variance were analyzed by one-way analysis of variance (ANOVA) and Bonferroni's multiple comparison test. Groups with unequal variance were analyzed by Welch's one-way analysis of variance (Welch's ANOVA) and Tamhane's T2 multiple comparison test. Non-normally distributed data were expressed as m (Q1-Q3) [median (25-75 percentile)] and compared by Kruskal Wallis test and Dunn's multiple comparison test. p < 0.05 was the significance threshold. The collected clinical data were expressed by Violin Plots and analyzed as described above. Chi-square test and Fisher's exact test were used to compare frequency data.

[0108] Experimental results:

[0109] 3.1 Control vs IgA vs IgG vs FLC biomarkers

[0110] Figure 7A is the statistical analysis of the Raman characteristic peak positions with VIP>0.5 of the four models: control vs IgA vs IgG vs FLC, control vs IgA, control vs IgG and control vs FLC.

[0111] Figure 7 B-7D shows the three peripheral blood biochemical indicators of total protein (TP), glucose, and high-density lipoprotein (HDL) in the control group, IgA group, IgG group, and FLC group. Statistical analysis shows that the peak intensity of the representative protein in the control group is higher than that in the MM groups, including 621, 643, 759, 1603, and 1616 cm -1 There were significant statistical differences compared with the IgA, IgG and FLC groups (P<0.01), and were consistent with the serological results ( Figure 7 B). The peak intensity of nucleic acid in the control group was higher than that in the MM group, with 1579 cm -1 There was a significant statistical difference compared with the MM groups (P<0.001), 726cm -1 There was a significant statistical difference compared with the FLC group (P<0.05), 781cm -1 There was a significant statistical difference compared with the IgA and IgG groups (P<0.05). The control group represents the peak position of β-carotene (957cm -1 ) was significantly higher than that of the MM groups (P<0.001).

[0112] The peak intensities of the control group, which represented carbohydrates (920, 1123 cm-1) and collagen (1345 cm-1), were lower than those of the MM group. -1 There were significant statistical differences compared with the MM groups ( Figure 7 A), consistent with the serological results ( Figure 7 C). The intensities of the peaks representing lipids in the control group (1437, 1443, 1446 cm-1) were higher than those in the MM group, which was consistent with the serological results ( Figure 7 D).

[0113] 3.2 Control group vs IgA group

[0114] Figure 8 A is the statistical analysis of the Raman characteristic peak positions with VIP>0.5 of the four models: control group vs IgA group, control group vs IgA-λ, control group vs IgA-κ, and IgA-λ vs IgA-κ.

[0115] Figure 8B-8D shows the peripheral blood biochemical indicators of TP, glucose and HDL in the control group, IgA-λ and IgA-κ groups. Statistical analysis shows that the peak intensity of the representative protein in the control group is higher than that in the IgA group, including 621, 643, 759, 1603 cm -1 There was a significant statistical difference compared with the IgA group (P<0.05), and it was consistent with the serological results ( Figure 8 B). The peak intensity of nucleic acid in the control group was higher than that in the IgA group, with 1579 cm -1 There was a significant statistical difference compared with the IgA groups (P<0.01), 781cm -1 There was a significant statistical difference compared with the IgA-λ group (P<0.05). The control group represents the peak position of β-carotene (957 cm -1 ) was significantly higher than that of the IgA-λ group (P<0.05).

[0116] The control group represents sugars (920, 1123 cm -1 ) and collagen (1345cm -1 ) was significantly lower than that of the IgA-κ group (P<0.05), which was consistent with the serological results ( Figure 8 C). The control group represents the peak positions of lipids (1437, 1443, 1446 cm -1 ) was significantly higher than that of the IgA group (P<0.01), which was consistent with the serological results ( Figure 8 D).

[0117] 3.3 Control group vs IgG group

[0118] Fig. 9 A is the statistical analysis of the Raman characteristic peak positions with VIP>0.5 of the four models: control group vs IgG group, control group vs IgG-λ, control group vs IgG-κ, and IgG-λ vs IgG-κ. Fig. 9 B-9D shows the peripheral blood biochemical indicators of TP, glucose and HDL in the control group, IgG-λ and IgG-κ groups. Statistical analysis shows that the peak intensity of the representative protein in the control group is higher than that in the IgG group, including 643, 759, 1603 and 1616 cm -1 There was a significant statistical difference compared with the IgG group (P<0.05), and it was consistent with the serological results ( Fig. 9 B). The peak intensity of nucleic acid in the control group was higher than that in the IgG group, with 1579 cm -1 There was a significant statistical difference compared with the IgG groups (P<0.01), 781cm -1There was a significant statistical difference compared with the IgG-κ group (P<0.05). The control group represents the peak position of β-carotene (957cm -1 ) was significantly higher than that of the IgG-κ group (P<0.001).

[0119] The control group represents sugars (920, 1123 cm -1 ) had a lower peak intensity than the IgG group, which was consistent with the serological results ( Fig. 9 C). The control group represents the peak positions of lipids (1437, 1443, 1446 cm -1 ) was significantly higher than that of the IgG group (P<0.05), which was consistent with the serological results ( Fig. 9 D). The control group represents collagen (1345 cm -1 ) was significantly lower than that of the IgG group (P<0.05).

[0120] 3.4 Control group vs FLC group

[0121] Fig.10 A is the statistical analysis of the Raman characteristic peak positions with VIP>0.5 of the four models: control vs FLC group, control vs FLC-λ, control vsκ, and FLC-λ vs FLC-κ. Fig.10 B-10D shows the peripheral blood biochemical indicators of TP, glucose and HDL in the control, FLC-λ and FLC-κ groups. Statistical analysis shows that the peak intensity of the representative protein in the control group is higher than that in the FLC group, including 621, 643, 759, 1603, 1616 cm -1 There was a significant statistical difference compared with the FLC group (P<0.05), and it was consistent with the serological results ( Fig.10 B). The peak intensity of nucleic acid in the control group was higher than that in the FLC group, with 1579 cm -1 There were significant statistical differences in the comparison with the FLC groups (P<0.05). The control group represented the peak position of β-carotene (957 cm -1 ) were significantly higher than those in the FLC group (P<0.001). The control group represented sugars (920, 1123 cm -1 ) and collagen (1345cm -1 ) was significantly lower than that of the FLC-λ group (P<0.001), which was consistent with the serological results ( Fig.10 C). The control group represents the peak positions of lipids (1443, 1446 cm -1 ) was significantly higher than that in the FLC-λ group (P<0.01), which was consistent with the serological results ( Fig.10 D).

[0122] Based on the above experiments, through the application of Raman spectroscopy (RS) technology, this patent discovered specific Raman peak position changes in serum related to MM. These peak position changes not only help to distinguish MM patients from healthy controls, but also provide possibilities for early diagnosis, typing and treatment response monitoring of MM.

[0123] References

[0124] [1]Belouni R,Allam I,Cherguelaine K,Berkani L,Belaid B,Berkouk Y,Nekkal S,Saidani M,Belhani M,Ghaffor M,Djidjik R.

[0125] Epidemiological and immunochemical parameters of monoclonal plasmacell dyscrasias of 2121cases in Algeria. Curr Res Transl Med.2020Apr;68(2):67-70.doi:10.1016 / j.retram.2019.11.003.Epub 2019Nov 21.PMID:31761679.

[0126] [2]Morris C,Drake M,Apperley J,Iacobelli S,van Biezen A,BjorkstrandB,Goldschmidt H,Harousseau JL,Morgan G,de Witte T,Niederwieser D,Gahrton G;Myeloma Subcommittee of Chronic Leukaemia Working Party of EBMT.Efficacy and outcome of autologous transplantation in rare myelomas.Haematologica.2010Dec;95(12):2126-33.doi:10.3324 / haematol.2010.022848.Epub 2010Oct22.PMID:20971818;PMCID:PMC2995572.

[0127] [3] Zhang L, Qi JY, Qi PJ, Wang YF, Zou DH, Yao HJ, An G, Yi SH, Li Q, Qiu LG. Comparison among immunologically different groups of 595 untreated multiple myeloma patients in northern China. Clin Lymphoma Myeloma Leuk. 2010 Jun;10(3):197 - 204. doi:10.3816 / CLML.2010.n.031. PMID:20511165.

[0128] [4] Landgren O, Kyle RA, Rajkumar SV. From myeloma precursor disease to multiple myeloma: new diagnostic concepts and opportunities for early intervention. Clin Cancer Res. 2011 Mar 15;17(6):1243 - 52. doi:10.1158 / 1078 - 0432.CCR - 10 - 1822. PMID:21411440; PMCID:PMC5901666. [5] Rosenbaum E, Marks D, Raza S. Diagnosis and management of neuropathies associated with plasma cell dyscrasias. Hematol Oncol. 2018 Feb;36(1):3 - 14. doi:10.1002 / hon.2417. Epub 2017 Apr 10. PMID:28397326.

[0129] [6]Leatham HW,Novoa R,Liedtke M,Kwong BY.Recurrent SubepidermalBlistering Dermatosis Heralding Disease Relapse in IgA Kappa MultipleMyeloma:Report of a Case and a Review of the Literature.Clin LymphomaMyelomaLeuk.2016 Jan;16(1):e1-5.doi:10.1016 / j.clml.2015.11.007.Epub 2015 Nov22.PMID:26708980.

Claims

1. Based on Raman spectroscopy, the Raman peaks representing nucleic acids in peripheral blood are detected at 1579, 726, and 781 cm -1 , representing protein Raman peaks 621, 643, 759, 1603, 1616 cm -1 , representing the Raman peaks of sugars at 920 and 1123 cm -1 , representing the Raman peak of collagen at 1345 cm -1 and the Raman peaks representing lipids at 1437, 1443, and 1446 cm -1 The use of the characteristic peak intensity level as a biomarker in the preparation of a product for predicting or detecting multiple myeloma.

2. The use according to claim 1, It is characterized in that The progression of the disease and the condition of the disease can be diagnosed and evaluated by detecting the peak intensity of the Raman peak biomarkers in the peripheral blood of the subjects and comparing it with the average value of the normal control.

3. The use according to any one of claims 1-2, It is characterized in that The multiple myeloma is of IgA, IgG or light chain (free light chain, FLC) type.

4. Based on Raman spectroscopy, the Raman peak representing β-carotene in peripheral blood is detected at 957 cm -1 , representing the Raman peak of nucleic acid at 781 cm -1 , representing sugars at 920 and 1123 cm -1 , representing 1345cm of collagen -1 The use of characteristic peak intensity level as a biomarker in the preparation of a product for typing IgA multiple myeloma, It is characterized in that The classification is to divide IgA into IgA-κ type and IgA-λ type.

5. The use according to claim 4, It is characterized in that The Raman peaks of the representative proteins of peripheral blood detected by Raman spectroscopy are 621, 643, 759, and 1603 cm -1 , representing the Raman peak of nucleic acid at 1579 cm -1 , and the peaks representing lipids are at 1437, 1443, and 1446 cm -1 The characteristic peak intensity level is used as a biomarker.

6. Based on Raman spectroscopy, the Raman peak representing β-carotene in peripheral blood is detected at 957 cm -1 , representing the Raman peak of nucleic acid at 781 cm -1 The use of characteristic peak intensity level as a biomarker in the preparation of a product for typing IgG type multiple myeloma, It is characterized in that The typing is to divide IgA type into IgG-κ type and IgG-λ type.

7. The use according to claim 6, It is characterized in that The Raman peaks of 643, 759, 1603, and 1616 cm representing proteins in peripheral blood were detected based on Raman spectroscopy. -1 , representing the Raman peak of nucleic acid at 1579 cm -1 , and the peaks representing lipids are at 1437, 1443, and 1446 cm -1 The characteristic peak intensity level is used as a biomarker.

8. Based on Raman spectroscopy, the Raman peaks representing lipids in peripheral blood are detected at 1443 and 1446 cm -1 , representing sugars at 920 and 1123 cm -1 , representing 1345cm of collagen -1 The use of characteristic peak intensity level as a biomarker in the preparation of a product for typing FLC type multiple myeloma, It is characterized in that The classification is to divide IgA types into FLC-κ type and FLC-λ type.

9. The use according to claim 8, It is characterized in that The Raman peaks of representative proteins in peripheral blood detected by Raman spectroscopy are 621, 643, 759, 1603, and 1616 cm -1 , representing the Raman peak of nucleic acid at 1579 cm -1 , and represents the β-carotene peak at 957 cm -1 The characteristic peak intensity level is used as a biomarker.

10. The use according to any one of claims 1 to 9, It is characterized in that The detection conditions of Raman spectroscopy are as follows: 785nm laser as excitation light, 40x objective lens, output power of 40mW, ×40 0.75NA Nikon lens used in the shooting process, the spot diameter range of about 2×2μm on the sample receives the laser beam with an output power of 40mW, the single integration time is 250s, the number of integrations is one, and the measurement range is 600-1800cm -1 , 5-10 sites were measured in each group, with a resolution of 1cm -1 .

Images