Two-dimensional magnetophoresis separation and detection method and system based on weight distribution

Abstract

The application discloses a two-dimensional magnetophoresis separation and detection method and system based on weight distribution, and belongs to the field of biochemical analysis. Based on the designed microfluidic chip, the physical property difference between the non-magnetic material and the magnetic material is utilized, the motion difference thereof in the magnetic field is taken as the distinguishing feature, deep-level signal exploration and analysis based on microbead quantification are realized, the number of the compound after the immune reaction and the motion difference in the magnetic field environment are utilized, force analysis calculation and physical field simulation are carried out, different weights are given to the classified signal probes, the signal strength is more accurately revealed, the two-dimensional magnetophoresis analysis method is combined, and the sensitivity and resolution capability of the analysis are effectively enhanced. In addition, the reaction time of the method is short, the operation steps are few, the multiple modes are optional, the scene applicability is strong, and the circulation use can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to a two-dimensional magnetophoretic separation and detection method and system based on weight distribution, belonging to the field of biochemical analysis. BACKGROUND

[0002] The biomarker strategy is essential for achieving high sensitivity detection and quantitative analysis in immunoassay, and is also an important tool for studying molecular interactions. When there is a specific reaction in immunoassay that meets the signal generation, the signal probe can significantly amplify the signal and be read out in a readable form by a specific instrument. This mode constitutes the basis of immunoassay. The principle of this method is that the content of the target is in a certain quantitative relationship with the signal probe, so it is widely used in the field of biomolecular analysis such as protein, nucleic acid analysis, pathogenic bacteria detection, drug residue, etc. The current labeling strategy mainly includes enzyme labeling, fluorescent labeling and microparticle labeling. Among them, enzyme-linked immunosorbent assay based on enzyme labeling is a classic immunoassay technology. It mainly relies on the specific capture of antibody molecules to the target, and then detects through enzyme amplification steps. However, this method often needs multiple washing and time-consuming operations, and the sensitivity is limited; chemiluminescence immunoassay based on enzyme-linked immunoassay improves the sensitivity and has been widely used in clinical diagnosis and other fields, but it still needs multiple washing steps, which limits its application; the immunoassay based on fluorescent labeling can achieve high sensitivity detection, and even single molecule level detection, so most digital immunoassay modes are based on this type of labeling. However, this method involves a complex signal processing system, long detection time, poor stability of fluorescent probes, and is not cost-effective, making it difficult to be widely applied; the immunoassay based on microparticle labeling has various forms, such as plasma labeling that can achieve high sensitivity detection, microbead labeling that is simple to operate, colloidal gold immunoassay method that is widely used for on-site detection, etc., but it is still difficult to have comprehensive advantages in detection sensitivity, analysis time, operation complexity and detection portability.

[0003] Magnetic particle labeling is a kind of microparticle labeling, which has the property of directional movement in a magnetic field. In immunoassay, it is usually used in combination with a magnetic field to realize target sorting, which can greatly reduce manual operation and shorten detection time, and has been widely used in chemiluminescence detection. It can also be combined with a portable optical imaging system to facilitate the development of a rapid on-site detection mode. However, the sensitivity of this labeling analysis method is relatively low. For example, the applicant's previous application for an analysis method based on magnetic separation for simultaneous detection of multiple target substances (CN 115047177A), a very important reason for low sensitivity is the signal processing system. For example, in the commonly used double-antibody sandwich immunoassay mode, non-magnetic materials (such as polystyrene microspheres) and magnetic materials are usually used in combination, and the two materials are labeled with two recognition molecules, respectively. When the immune sandwich structure is formed, a complex of the two materials is formed, and one complex (such as a polystyrene microsphere) is usually considered as one signal. However, one complex often contains multiple magnetic particles, so the content information of the target substance carried by each complex is different, but after uniform processing, it is only considered as one microsphere signal, which will cause signal loss, which limits the improvement of the sensitivity of the method. The above-mentioned traditional biological labeling method is one-dimensional, that is, based on the quantitative relationship between the content of the target substance and the number of probes. When the content of the target substance is low, multiple target substances may correspond to one probe signal, resulting in that at low concentration, it is not possible to distinguish the one-to-one correspondence between the probe and the content of the target substance, leading to signal loss. Therefore, the traditional biological labeling method has the disadvantage of measurement uncertainty in the case of low concentration of target substance. Therefore, a multi-step washing process is required to remove the unreacted probes, so the entire reaction is a heterogeneous reaction, and the operation is relatively complex.

[0004] In addition, in order to simplify the steps and avoid multi-step washing, homogeneous wash-free biological labeling analysis methods have also attracted widespread attention. The most typical ones are fluorescence resonance energy transfer or distance-based chemiluminescence methods. These methods also label antibody or antigen molecules on micro-nano spheres, and through specific recognition of antibody-antigen, the distance is pulled in, thereby causing a change in the signal. This homogeneous wash-free method greatly improves the simplicity of operation and shortens the detection time. Light-activated chemiluminescence detection is a typical homogeneous immunoassay method. However, because of its relatively large background signal, the sensitivity and linear range of the method are relatively narrow, which limits its practical application. SUMMARY

[0005] To solve the above problems, the application provides a two-dimensional magnetophoresis separation and detection method and system with high sensitivity, simple operation (homogeneous, one-step, and no washing) and multi-scene applicability. The application selects non-magnetic polymer insulating microspheres and magnetic nanoparticles as a composite signal probe, fully designs the multifunctionality of the magnetic nanoparticles, integrates a microfluidic chip, one-step immune reaction, weight analysis, and signal reading technology based on deep learning. The method can expand various analysis modes, realizes one-step washing-free immune reaction with high sensitivity in multiple scenes, and specific technical solutions are as follows:

[0006] The first object of the application is to provide a two-dimensional magnetophoresis separation and detection method, comprising:

[0007] Step 1: coupling a biological recognition molecule A of a target to be measured on the surface of a non-magnetic polymer insulating microsphere carrier with a certain particle size and / or color;

[0008] Step 2: coupling a biological recognition molecule B of the target to be measured on the surface of a magnetic nanoparticle, wherein the biological recognition molecule A matches the biological recognition molecule B;

[0009] Step 3: mixing a non-magnetic polymer microsphere carrier coupled with the biological recognition molecule A, a magnetic nanoparticle coupled with the biological recognition molecule B, and a liquid containing the target to be measured to perform a homogeneous reaction, to obtain a mixed liquid containing reaction complexes with different magnetic susceptibilities;

[0010] Step 4: injecting the mixed liquid obtained in step 4 into a microfluidic chip, and under the action of a magnetic field in a working area of the microfluidic chip, performing magnetophoresis separation on the reaction complexes with different magnetic susceptibilities, wherein the microfluidic chip comprises linearly arranged microtubes for anchoring the reaction complexes with different magnetic susceptibilities;

[0011] Step 5: collecting an image of the working area of the microfluidic chip, wherein the image covers each microtube after magnetophoresis separation;

[0012] Step 6: based on the position of each microtube, using a finite element analysis software to simulate the physical field where the microfluidic chip is located, obtaining the stress value of the reaction complex in each microtube, and converting the stress value into the weight of each microtube;

[0013] Step 7: using a machine vision recognition model to segment the accumulation area of the microsphere probe in each microtube, and calculating the pixel intensity;

[0014] Step 8: based on the weight obtained in step 6 and the pixel intensity identified in step 7, calculating the weight-pixel intensity of each microtube, and the calculation method of the weight-pixel intensity is: weight x identified pixel intensity, and adding the weight-pixel intensities of all microtubes to obtain the total pixel intensity;

[0015] Step 9: Based on the total pixel intensity, the quantitative detection of the concentration of the target object to be detected is realized in combination with the fitting curve between the total pixel intensity and the concentration of the target object to be detected.

[0016] Optionally, the step 6 analyzes the Stokes drag force provided by the fluid and the magnetic force provided by the magnetic field on the reaction complex.

[0017] Optionally, the non-magnetic polymer insulating microsphere carrier includes but is not limited to polystyrene microspheres, silica microspheres, polybutadiene microspheres, and polyisoprene microspheres.

[0018] Optionally, the particle size of the magnetic nanoparticles is 30-1000 nm.

[0019] Optionally, the particle size of the non-magnetic polymer microsphere carrier is 1-50 μm.

[0020] Optionally, the target object to be detected includes but is not limited to viruses, antibodies, biomarkers, antibiotic molecules, pesticide molecules, veterinary drug molecules, biological toxins, bacteria, and detection objects of various matrix components, such as serum, milk, beverages, or others.

[0021] Optionally, the biological recognition molecule A and the biological recognition molecule B include capture antibodies and detection antibodies, detection antibodies and capture antibodies, antibodies and antigens, antigens and antibodies, DNA capture probes and DNA detection probes, and DNA detection probes and DNA capture probes.

[0022] Optionally, the machine vision recognition model is a YOLOv7 image segmentation model.

[0023] Optionally, the color of the non-magnetic polymer insulating microsphere carrier includes but is not limited to purple, red, orange, black, blue, green, yellow, or white.

[0024] A second object of the present application is to provide a two-dimensional magnetic phoresis separation and detection system, comprising a microfluidic chip, an optical imaging device, and a computing device.

[0025] The microfluidic chip comprises a stable region W1, a correction region W2, and a working region W3 connected in sequence and in conduction, wherein the correction region W2 and the working region W3 are located in a magnetic field affected region W4, and reaction complexes with different magnetic susceptibilities in a mixed solution are anchored by a linearly arranged microtube array in the working region W3 under the action of a magnetic field.

[0026] The optical imaging device is used to collect images of the working region W3.

[0027] The computing device comprises:

[0028] An image acquisition module configured to acquire a microtube array image collected by the optical imaging device;

[0029] A weight calculation module configured to simulate, based on the position of each microtube, a physical field in which the microfluidic chip is located by using finite element analysis software, obtain a stress value of a reaction complex in each microtube, and convert the stress value into a weight of each microtube;

[0030] An image segmentation module configured to segment a stacking area of a microsphere probe in each microtube by using a machine vision recognition model, and calculate a pixel intensity;

[0031] A pixel intensity calculation module configured to calculate a post-weight pixel intensity of each microtube based on the weight and the recognized pixel intensity, the calculation method of the post-weight pixel intensity being: weight * recognized pixel intensity, and adding the post-weight pixel intensities of all microtubes to obtain a total pixel intensity;

[0032] A quantitative detection module configured to realize quantitative detection of the concentration of the target object to be detected based on the total pixel intensity and in combination with a fitting curve between the total pixel intensity and the concentration of the target object to be detected.

[0033] Optionally, the optical imaging device includes but is not limited to a bright field microscope, a dark field microscope, a smart phone, and a portable lens.

[0034] Optionally, the microfluidic chip reaction counting plate substrate material is any one of quartz glass, polyethylene, and polyvinyl chloride; the counting cavity material bonded to the substrate is any one of HTL high-precision photosensitive resin, PDMS (polydimethylsiloxane), PMMA (polymethyl methacrylate), and quartz glass; and the microtube linear array shape is an arrayed polygon combination, but is not limited to such a shape.

[0035] Optionally, the number of microtubes in the microfluidic chip linear array is 2 or more, including but not limited to 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, and the like.

[0036] Optionally, the magnetic field can be generated by a permanent magnetic material or an electromagnet, and the permanent magnetic material includes neodymium, iron, boron, cobalt, nickel, rare earth elements, and alloys thereof.

[0037] Optionally, the magnetic field strength is 0.01 T or more, including but not limited to 0.01 T, 0.02 T, 0.03 T, 0.04 T, 0.05 T, 0.06 T, 0.07 T, 0.08 T, 0.09 T, 0.10 T, and the like.

[0038] The working principle of the application is as follows:

[0039] The present application selects a non-magnetic high polymer insulating microsphere of a particle size and / or color as a carrier and signal probe, a magnetic nano-particle as a multifunctional element, modifies a biological recognition molecule (capture antibody, complete antigen, DNA probe, etc.) corresponding to the target to be detected on the surface of the non-magnetic high polymer insulating microsphere, and modifies a corresponding detection antibody or antibody matching biological recognition molecule on the surface of the magnetic nano-particle. Taking a sandwich immunoassay as an example (corresponding to the capture antibody and the detection antibody), as shown in Figure 1 When the target exists, the capture antibody modified on the surface of the non-magnetic high polymer insulating microsphere captures the target on the surface of the non-magnetic high polymer insulating microsphere, and the magnetic nano-particle labeled with the detection antibody, to form a non-magnetic high polymer insulating microsphere carrier-capture antibody-target-detection antibody-magnetic nano-particle complex of a particle size and / or color.

[0040] The target content will affect the complex in two dimensions. First, from a macroscopic point of view, the total number of the complex of the structure of the non-magnetic high polymer insulating microsphere carrier-capture antibody-target-detection antibody-magnetic nano-particle of a particle size and / or color changes; second, from a microscopic point of view, due to the current microsphere surface modification technology and the difference in contact area between the non-magnetic high polymer insulating microsphere and the magnetic nano-particle, the signal probe and the multifunctional element in the complex are not a simple one-to-one relationship, that is, the number of magnetic nano-particles contained in each complex of the structure of the non-magnetic high polymer insulating microsphere carrier-capture antibody-target-detection antibody-magnetic nano-particle of a particle size and / or color is different. After the immunoassay, the content of the target to be detected is proportional to the number of signal probes of the non-magnetic high polymer insulating microsphere of different particle sizes and / or colors in the complex, and is proportional to the number of magnetic nano-particles in the complex.

[0041] When the above complex is injected into the microfluidic chip by the fluid pushing device, it flows through the micro-pipeline linear array at a predetermined speed through the inlet I1, as shown in Figure 2As shown, the microfluidic chip is divided into three main regions: a stabilization region W1, a correction region W2 and a working region W3, where W2 and W3 are in the magnetic field influence region W4. W1 is less affected by the magnetic field, where the micro-particles loaded into the chip reach a relatively stable state. W2 and W3 are affected by the magnetic field, where the complexes entering these regions experience magnetophoretic movement in the presence of an external magnetic field. W2 is used to initially adjust the position of the complexes with magnetic susceptibility. It guides them close to the edge of the smooth channel, ensuring that these complexes enter W3 with the same position in the y direction. Here, the complexes with different magnetic susceptibility are anchored by the microtubule array due to magnetic force, while the polymer microspheres without magnetic susceptibility are discharged from the outlet (I2). In the linear array of microtubules, the main forces for the two-dimensional differential sorting of complexes are the Stokes drag force provided by the fluid and the magnetic force provided by the magnetic field. Compared with the non-magnetic polymer insulating microspheres, the external magnetic field of the microfluidic chip will make the complexes with magnetic susceptibility produce obvious magnetically driven motion. Therefore, the complexes move towards the magnet and are finally anchored on the linear array of microtubules, while the non-magnetic polymer insulating microspheres that fail to capture the target continue their linear motion and are discharged.

[0042] The depth-guided design of the linear microtubules of the microfluidic chip makes the accumulation of complexes have a sensitive observable scale to reflect the one-dimensional number change of the complexes. In addition, the complexes affected by the magnetic field show different magnetophoretic trajectories according to their different magnetic susceptibility, in the W3 region of the microfluidic chip, see Figure 3 As shown, the allowed deflection displacement of the complex in the y direction is a certain value. When this displacement threshold is reached, it will be anchored by the microtubule. The complexes with different magnetic susceptibility need different time to achieve the equivalent displacement in the y direction under the influence of the magnetic force.

[0043] During this period, the Stokes drag force acting on the complex in the x direction will separate the displacement of the immune complex in the x direction. The Stokes drag force mainly provides the force for the movement of the complex in the x direction, which can be represented as where η is the viscosity of the fluid; r p is the radius of the complex; is the velocity of the fluid; is the velocity of the complex. The magnetic force provides the movement of the complex in the y direction, which can be represented as, where N is the number of magnetic nanoparticles combined with the polystyrene microspheres; Δχ is the magnetic susceptibility difference between the magnetic nanoparticles and the solution; μ0 is the magnetic permeability of free space; V m is the volume of the magnetic nanoparticles; is the magnetic field; is the del operator. The vector nature of the force allows independent consideration of forces in different directions.

[0044] The movement of the complex in the y direction can be analyzed by Newton's second law, Fy = a y M wherein F y is the resultant force in the y direction, M is the mass of the complex, a y is the acceleration of the complex in the y direction. F y is mainly provided by the magnetic force in the y direction and the Stokes drag force in the y direction.

[0045] In classical mechanics, the displacement of the complex in the y direction can be expressed as: where L y is a predetermined distance established in the y direction of the partitioned area, v py0 is the initial speed of the immune complex in the y direction, and t is the time required for the immune complex to reach the displacement in the y direction. Since the immune complex has no initial speed in the y direction after passing through W2, t is equal to 0, and the formula for the time required for the complex to produce a y direction length displacement can be expressed as: The x direction displacement of the complex with different magnetic loads can be expressed as: where υ px is the speed of the immune complex in the x direction under the action of the Stokes drag force, and the time interval allows the immune complex to produce a displacement difference in the x direction, thereby facilitating the sorting of complexes with different magnetic loads. The linear arrangement of microtubes can record and reflect these differences to reflect the two-dimensional trajectory changes of the complex.

[0046] The present application adopts an optical imaging system to read signals, as shown in Figure 4 , the microfluidic chip working area is placed at the imaging device, and after imaging, the image is transmitted to a computer or a mobile phone. The visual recognition model based on YOLOv7 encodes the microtubes in space, and according to the weights allocated by pre-physical field simulation and force analysis, the information of each microtube in the linear array is accurately positioned and read, and finally the information about the target concentration is output.

[0047] The present application uses non-magnetic high polymer insulating microspheres with different particle sizes and / or colors as carriers and signal probes, couples biological recognition molecules on the surface of the non-magnetic high polymer insulating microspheres, specifically binds to the target object through sandwich, competition, DNA hybridization, etc., introduces magnetic nanoparticles, can quickly mark the target object and convert the target object content into the magnetic susceptibility of the complex, after two-dimensional magnetophoresis analysis, realizes one-step, no-washing, high-sensitivity and rapid detection of the target object.

[0048] The present application has the following beneficial effects:

[0049] (1) Based on the signal probe two-dimensional magnetophoresis separation, the signal analysis depth of the microbead sensing is expanded, and the signal is analyzed from the two-dimensional angle, which significantly improves the sensitivity and resolution of the analysis method. Compared with the previous quantitative immunization strategy based on microbeads, the quantity of the complex after the immunization reaction is directly quantified, which can achieve relatively sensitive detection, but ignores the target information carried by the complex at the microscopic level, and the signal is analyzed according to the uniform quantitative standard, which causes the loss of signal in principle. The method proposed in the application utilizes the physical property difference between non-magnetic materials and magnetic materials, and the motion difference in the magnetic field as the distinguishing feature, realizes the deep signal mining and analysis based on the microbead quantification, and realizes the two-dimensional magnetophoresis analysis through the quantity and motion difference of the complex after the immunization reaction in the magnetic field environment, thereby significantly improving the sensitivity and resolution of the method.

[0050] (2) Short reaction time and few operation steps. The magnetic nanoparticles are designed as multifunctional elements, and the abilities of signal labeling, magnetic separation, signal interpretation and other related functions are integrated. These properties enable the method to realize homogeneous one-step detection without washing, and the steps of repeated washing and magnetic separation of immune complexes are not required. In the application, the number of polymer microsphere signal probes in the complex is directly related to the content of the target, and the whole signal conversion is completed in one step without additional signal amplification steps.

[0051] (3) Multi-mode optional, strong scene applicability, and can be recycled. The microfluidic chip designed in the application is a linear arrangement type with a relatively simple structure, and has stronger compatibility with the pushing device. It can not only realize the delivery of the reaction liquid to the inside of the microfluidic chip through an external injection pump, but also can deliver the liquid to the inside of the microfluidic chip through a simple capillary structure trigger structure through capillary force. This makes the method not only suitable for laboratory, but also suitable for various scene detection requirements such as field detection. In addition, the microfluidic chip can be cleaned by replacing the position of the magnet, and can be recycled multiple times.

[0052] (4) Personalized signal processing scheme cooperates to enhance the detection performance. The two-dimensional magnetophoresis separation principle proposed in the application is matched with a personalized signal processing scheme, different weights are given to the classified signal probes through force analysis calculation and physical field simulation, and the signal strength is more accurately revealed, which cooperates with the two-dimensional magnetophoresis analysis method to enhance the sensitivity and resolution of the analysis. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0054] Figure 1 The biochemical reaction principle flow chart for detecting the marker in the sandwich immunoassay in the application.

[0055] Figure 2 The microfluidic chip design chart in the application.

[0056] Figure 3 The principle diagram for two-dimensional magnetophoresis separation of the complex in the working area of the microfluidic chip in the application.

[0057] Figure 4 The sorting and identification diagram of the microfluidic chip in the application.

[0058] Figure 5 The diagram for the machine vision to recognize the microtube information according to the matching weight in the method of Example 1.

[0059] Figure 6 The biochemical reaction result chart and analysis chart for detecting interleukin 6 in serum in the laboratory mode of the method of Example 1.

[0060] Figure 7 The comparison chart of the analysis results of the weight analysis method and the non-weight analysis method in Example 1.

[0061] Figure 8 The standard curve comparison chart of the method of the application and the standard chemiluminescence method for detecting interleukin 6 in serum in Example 1.

[0062] Figure 9 The standard curve and sensitivity comparison chart of the method of the application and the one-dimensional analysis strategy based on the number of microspheres for detecting interleukin 6 in serum in Example 1.

[0063] Figure 10 The result chart of the method of the application for detecting chloramphenicol in the general mode in Example 2.

[0064] Figure 11 The result chart of the method of the application for studying the interaction between mannose and bacteria in the general mode in Example 3.

[0065] Figure 12 The microfluidic built-in self-driven capillary pump structure diagram in the field mode of the method of the application in Example 4.

[0066] Figure 13 The result chart and analysis chart of the method of the application for detecting the N protein of the new coronavirus in the field mode in Example 4.

[0067] Figure 2 I1: microfluidic chip inlet, W1: stabilization zone, W2: correction zone, W3: working zone, W4: magnetic field influence zone, I2: microfluidic chip outlet. DETAILED DESCRIPTION

[0068] The application will be described in detail below with reference to specific embodiments.

[0069] The reagents used in the experiment and the related terms are explained as follows:

[0070] Carboxylated polystyrene microspheres: 5 μm polystyrene microspheres, color including white and red; purchased from Bangs Laboratories, Inc.

[0071] Carboxylated superparamagnetic nanoparticles: 150 nm; purchased from Ocean Nanotech, LLC.

[0072] Interleukin-6 (IL-6), recombinant N protein of new coronavirus, and the corresponding capture antibody and detection antibody were purchased from abcam. E. coli JM109 (ATCC 53323), S. typhimurium (ATCC 14028), and S. aureus (ATCC 29213) were purchased from American Type Culture Collection. Chloramphenicol (CAP), chloramphenicol complete antigen (CAP-BSA), and chloramphenicol antibody were purchased from Bioss.

[0073] 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDC), N-hydroxysulfosuccinimide ester (sulfo-NHS), and 4-aminophenyl-α-D-mannopyranoside (98%) were purchased from Shanghai Aladdin Biochem Technology Co., Ltd. Tween-20 and bovine serum albumin (BSA) were purchased from Amresco.

[0074] PBS buffer (10 mM, pH = 7.4): 8.00 g of NaCl, 0.20 g of KCl, 0.20 g of KH2PO4, and 2.90 g of Na2HPO4·12H2O were taken in a 1000 mL volumetric flask, and then the volume was made up with water and shaken well.

[0075] MES buffer (0.1 M, pH = 6.0): 21.325 g of MES was dissolved in deionized water to make up to 1000 mL as solution A; 4 g of NaOH was dissolved in deionized water to make up to 1000 mL as solution B; 1000 mL of solution A and 400 mL of solution B were mixed and shaken well.

[0076] PBST, MEST: 0.5 mL of Tween-20 was added to the prepared 1000 mL of PBS or MES buffer, and shaken well.

[0077] Example 1: Quantitative detection of the inflammatory marker interleukin-6 (IL-6) using a sandwich assay in a laboratory mode

[0078] (1) Preparation of the reagents for the reaction

[0079] 5 pm red carboxyl functionalized polystyrene microspheres were selected as the polystyrene microsphere signal probe-capture antibody conjugate carrier, and a 5 pm red polystyrene microsphere signal probe-IL-6 capture antibody conjugate complex was prepared. 150 nm carboxyl functionalized magnetic particles were selected as the separation carrier, and IL-6 detection antibodies were modified on the surface thereof. The preparation method of the microsphere conjugate is a conventional method known in the art, and is specifically as follows:

[0080] Preparation of the polystyrene microsphere carrier-IL-6 capture antibody conjugate complex: 1 mg of carboxyl polystyrene microspheres (5 pm, red, 50 mg / mL) was washed twice with MEST buffer (pH = 6.0 containing 0.05% Tween 20), resuspended with MES (pH = 6.0) buffer, and then 30 pL of EDC (10 mg / mL) and 30 pL of NHS (10 mg / mL) were added to 500 mL. After mixing, the mixture was activated at room temperature for 15 min. After activation, the mixture was washed twice with PBS (pH = 7.4) buffer and centrifuged, resuspended with 500 ml of PBS buffer, and then 50 pg of IL-6 capture antibody was added. The mixture was shaken at 37°C for 2-4 h. After the reaction was completed, the mixture was blocked with 1% BSA solution (diluted with PBS) for 30 min. After blocking was completed, the mixture was washed four times with PBST (pH = 7.4 containing 0.05% Tween 20) buffer. The polystyrene microsphere-IL-6 capture antibody conjugate was resuspended with PBS buffer (containing 0.5% BSA) and stored at 4°C for use.

[0081] Preparation of the magnetic nanoparticle-IL-6 detection antibody conjugate complex: 1 mg of carboxyl magnetic nanoparticles (150 nm, 10 mg / ml) was washed twice with MEST buffer (pH = 6.0 containing 0.05% Tween 20), resuspended with MES (pH = 6.0) buffer, and then 30 pL of EDC (10 mg / mL) and 30 pL of NHS (10 mg / mL) were added to 500 mL. After mixing, the mixture was activated at room temperature for 15 min. After activation, the mixture was washed twice with PBS (pH = 7.4) buffer and magnetically separated, resuspended with 500 ml of PBS buffer, and then 20 pg of IL-6 capture antibody was added. The mixture was shaken at 37°C for 2-4 h. After the reaction was completed, the mixture was blocked with 1% BSA solution (diluted with PBS) for 30 min. After blocking was completed, the mixture was washed four times with PBST (pH = 7.4 containing 0.05% Tween 20) buffer. The magnetic particle-IL-6 capture antibody conjugate complex was resuspended with PBS buffer (containing 0.5% BSA) and stored at 4°C for use.

[0082] (2) Homogeneous wash-free double antibody sandwich immune reaction

[0083] 5 μL of 2 mg / mL polystyrene microspheres signal probe coupled with IL-6 capture antibody, 6 μL of 1 mg / mL magnetic nanoparticles coupled with IL-6 detection antibody and 10 μL of target sample to be tested were mixed, after 25 min of reaction at room temperature, 10 μL of the reaction solution was pumped into a microfluidic chip containing 50 microtubes at a rate of 1.5 μL / min, at the same time, the working area of the chip started to work, and after the pumping was completed, the chip synchronous sorting ended.

[0084] (3) Imaging device photographing

[0085] The working microfluidic chip was placed in the designated position of the optical microscope imaging, the imaging objective lens (10X), the imaging eyepiece (10X) and the CMOS industrial camera were placed directly above the microfluidic chip, the imaging objective lens position and the imaging light source brightness were adjusted until a clear and visible linear microtube accumulated microsphere image appeared on the CMOS imaging software and the imaging field covered the entire working area; then, the working area was photographed by the CMOS industrial camera, and the imaging image was transmitted to the computer for saving through the gigabit data switch.

[0086] (4) Microtube weight distribution

[0087] The microfluidic chip layout of 50 microtube linear array was adopted, the motion of the complex was further studied to distribute the weight of the linear array microtube, and the main factors affecting the position of the complex in the linear array channel were the Stokes drag force and the magnetic force.

[0088] The Stokes drag force mainly provides the force for the motion of the complex in the x direction, which can be expressed as where η is the viscosity of the fluid; r p is the radius of the complex; is the velocity of the fluid; is the velocity of the complex.

[0089] The magnetic force provides the motion of the complex in the y direction, which can be expressed as where N is the number of magnetic nanoparticles combined with the polystyrene microspheres; Δχ is the magnetic susceptibility difference between the magnetic nanoparticles and the solution; μ0 is the magnetic permeability of free space; V m is the volume of the magnetic nanoparticles; is the magnetic field; is the del operator.

[0090] The motion of the complex in the y direction can be analyzed by Newton's second law, F y = a y M where F y is the resultant force in the y direction, M is the mass of the complex, and ay is the acceleration of the complex in the y direction. F y is mainly provided by the magnetic force in the y direction and the Stokes drag force in the y direction.

[0091] In classical mechanics, the displacement of the complex in the y direction can be expressed as: where L y is a predetermined distance established in the y direction of the partitioned region, v py0 is the initial velocity of the immune complex in the y direction, t is the time required for the immune complex to reach the displacement in the y direction, since the immune complex has no initial velocity in the y direction after passing through W2, t is equal to 0, and the formula for the time required for the complex to produce a y direction length displacement can be expressed as: The x direction displacement of the complex with different magnetic loads can be expressed as: where v px is the velocity of the immune complex in the x direction under the action of the Stokes drag force, and the time interval allows the immune complex to produce a displacement difference in the x direction, thereby promoting the sorting of complexes with different magnetic loads, and the linear arrangement of microtubes can record and reflect these differences to reflect the two-dimensional trajectory changes of the complex.

[0092] Based on the above analysis principle, the microtubule linear array of the microfluidic chip is introduced into the finite element analysis software COMSOL, the physical field where the microfluidic chip is located is simulated by setting the microsphere magnetic susceptibility, fluid velocity, magnetic field strength and other information, and the COMSOL finally outputs the landing point parameters. The landing point parameters can be horizontal and vertical coordinate data, the magnetic susceptibility of different complex microspheres is different, the obtained landing point parameters are different, and the weight value of each microtube is obtained after normalization processing, which reflects the ratio between the force values of the microtubes. It can be used to measure the concentration information of the microsphere probe in the microtube. The weight distribution of each microtube in the embodiment is shown in Table 1.

[0093] Table 1 Distribution weight value of different microtubes corresponding to chip microtube array

[0094]

[0095]

[0096] (5) Image processing and machine vision recognition

[0097] The segmentation of the stacked probe region in the chip linear array and the coordinate calibration between the microtubes are processed by the channel quantitative coding algorithm (CQCA), which involves using the YOLOv7 image segmentation model to segment the stacked probe region and the coordinate calibration between the microtubes based on the computer vision (CV) - image positioning algorithm.

[0098] In the stacked probe area segmentation model, first, images of different concentrations are obtained to establish an image segmentation dataset. Data augmentation algorithms are used to increase the robustness of the model, and the dataset is divided into a training set and a validation set. Second, the model based on the improved YOLOv7 image segmentation is used to segment the microsphere probe accumulation area, and the conversion to pixel intensity is used to reflect the accumulation amount of the microsphere probe. The model uses pre-trained weights from Yolov7-seg.pt, trains for more than 300 epochs, with a batch size of 16, an initial learning rate of 0.01, and an input network image size of 512. Training is performed on a GPU, using an SGD optimizer and a multi-scale training method. During the prediction process, the image size is adjusted to meet the requirements of the long channel image.

[0099] In the coordinate calibration, the algorithm first binarizes the segmented image, uses the CV algorithm to correct the horizontal encoding of the segmentation mask image, and encodes 0 to 49 for the 50 microtubes, and calibrates each microtube. In addition, it also introduces additional restrictions. In the vertical direction, the areas with overlapping horizontal coordinate calibration in the shielded area are merged to eliminate the influence of discrete stacked probes. As shown in Figure 5 The calibrated mask image establishes a mapping relationship with the original image, calculates the pixel intensity of each microtube, and then retrieves the pixel intensity value according to the microtube code, which helps to comprehensively quantify the microspheres on different microtubes.

[0100] The development of the recognition model is based on Python 3.8, and is supported by algorithm packages such as PyTorch 3.0.1+cu11.8, NumPy 1.21.5, OpenCV 4.6.0, Pandas 1.2.4. The GPU model used for model training is NVIDIA GeForce RTX 3090Ti, and the CPU model is i9-12900K.

[0101] (6) Quantitative detection of IL-6 in human serum samples

[0102] After the image processing process, the YOLOv7-based recognition model will recognize the signal intensity of each microtube, i.e. the pixel intensity, according to the pre-calculated weights. The pixel intensity of each microtube is calculated as: the weight of this microtube x the pixel intensity value recognized by this microtube. Then, the total pixel intensity of all microtubes is summed to obtain the total pixel intensity after weighting. With the total pixel intensity after weighting as the vertical coordinate and the logarithmic value of the IL-6 concentration as the horizontal coordinate, a fitting curve between the two is obtained. In the actual detection scenario, after obtaining the total pixel intensity of the microtube after weighting, the quantitative detection of IL-6 can be performed according to the fitting curve.

[0103] The key parameters of the biochemical reactions involved in this embodiment are optimized, and the results of the IL-6 detection under the optimized conditions are as shown in Figure 6 The results show that the present method has a good linear relationship between the weighted pixel intensity and the IL-6 concentration in the range of 5 pg / mL to 100000 pg / mL, and r 2 = 0.99. At the same time, in order to reflect the advantages of the weight processing, for the same group of samples obtained in the experiment, the samples are analyzed by the weight and the non-weight analysis strategies respectively, as shown in Figure 7 The non-weight analysis processing scheme has limited resolution for low concentration and adjacent concentration samples, while the analysis scheme based on the weight of the present application has significant improvement in analysis sensitivity and resolution. Compared with the standard chemiluminescence detection method for detecting IL-6, as shown in Figure 8 The detection sensitivity of the present application is improved by nearly 50 times (the minimum detection limit of the present application is 630 fg / mL, and the chemiluminescence is 31.2 pg / mL), as shown in Figure 9 Compared with the one-dimensional analysis strategy based on the number of microspheres (the minimum detection limit is 14.2 pg / mL), the present application has a 23-fold improvement in detection sensitivity. The detection method used in the present application has shorter detection time, no need for multi-step cleaning, and wider linear range and higher sensitivity, which is crucial for user-friendly real sample biomarker detection.

[0104] Example 2, using a general mode, taking competitive immunity as an example, quantitative detection of chloramphenicol (CAP) residues

[0105] (1) Preparation of reaction reagents

[0106] 5 μm white carboxyl functionalized polystyrene microspheres were selected as the polystyrene microsphere signal probe-CAP-BSA conjugate carrier, 5 μm white polystyrene microsphere signal probe CAP-BSA conjugate complex was prepared, 150 nm carboxyl functionalized magnetic particles were selected as the separation carrier, and CAP antibody was modified on the surface thereof. The microsphere conjugate preparation method is a conventional method known in the art, and the specific process is as follows:

[0107] Preparation of polystyrene microsphere carrier-CAP-BSA conjugate complex: take 1 mg carboxyl polystyrene microspheres (5 μm, white, 50 mg / mL), wash twice with MEST buffer (pH = 6.0 containing 0.05% Tween 20), resuspend with MES (pH = 6.0) buffer, add 30 μL EDC (10 mg / mL) and 30 μL NHS (10 mg / mL) to 500 mL, mix well, and activate at room temperature for 15 min. After activation, wash twice with PBS (pH = 7.4) buffer, centrifugal separation, resuspend with 500 ml PBS buffer, add 50 μg CAP-BSA, and react at 37°C for 2-4 h; after reaction, block with 1% BSA solution (diluted with PBS) for 30 min, wash 4 times with PBST (pH = 7.4 containing 0.05% Tween 20) buffer after blocking, resuspend the polystyrene microsphere-CAP-BSA conjugate with PBS buffer (containing 0.5% BSA), and store at 4°C for use.

[0108] Preparation of magnetic nanoparticle-CAP antibody conjugate complex: take 1 mg carboxyl magnetic nanoparticles (150 nm, 10 mg / ml), wash twice with MEST buffer (pH = 6.0 containing 0.05% Tween 20), resuspend with MES (pH = 6.0) buffer, add 30 μL EDC (10 mg / mL) and 30 μL NHS (10 mg / mL) to 500 mL, mix well, and activate at room temperature for 15 min. After activation, wash twice with PBS (pH = 7.4) buffer, magnetic separation, resuspend with 500 ml PBS buffer, add 20 μg CAP antibody, and react at 37°C for 2-4 h; after reaction, block with 1% BSA solution (diluted with PBS) for 30 min, wash 4 times with PBST (pH = 7.4 containing 0.05% Tween 20) buffer after blocking, resuspend the magnetic particle-CAP antibody conjugate complex with PBS buffer (containing 0.5% BSA), and store at 4°C for use.

[0109] (2) Homogeneous cleaning-free competitive immune reaction

[0110] Mix 4 μL 2 mg / ml CAP-BSA conjugated polystyrene microsphere signal probe with 6 μL 1 mg / mL CAP antibody conjugated magnetic nanoparticle and 10 μL target sample to be detected, react at room temperature for 15 min, then pump 10 μL reaction solution into the microfluidic chip at 1.5 μL / min by microsyringe pump, at the same time, start the work of the chip, and after pumping, end the chip synchronous sorting.

[0111] (3) Imaging device takes a picture

[0112] The microfluidic chip working position is placed in the imaging designated position of the optical microscope, the imaging objective (4X), the imaging eyepiece (10X) and the CMOS industrial camera are placed right above the microfluidic chip, the imaging objective position and the imaging light source brightness are adjusted until the CMOS imaging software shows a clear linear microtubule accumulation microsphere image and the imaging field of view covers the whole working area; then, the working area is photographed by the CMOS industrial camera, and the imaging image is transmitted to the computer for saving through the gigabit data switch.

[0113] The microtubule weight distribution, image processing and machine vision recognition steps are consistent with Embodiment 1, and the CAP detection linear range is detected according to the above operation process, as shown in Figure 10 The results show that the present application has a good linear relationship between the weighted pixel intensity and the concentration of CAP in the range of 10-10 4.5 pg / mL, r 2 = 0.99.

[0114] Example 3: Study of the interaction between sugar and bacteria using a general mode

[0115] (1) Preparation of reaction reagents

[0116] 5 μm white carboxyl functionalized polystyrene microspheres are selected as the polystyrene microsphere signal probe-CAP-BSA conjugate carrier, 5 μm white polystyrene microsphere signal probe mannose conjugate complex is prepared, and 150 nm carboxyl functionalized magnetic particles are selected as the separation carrier, and the surface of the magnetic particles is modified with mannose. The microsphere conjugate preparation method is a conventional method known in the art, and the specific method is as follows:

[0117] Preparation of polystyrene microsphere carrier mannose conjugate complex: take 1 mg of carboxyl polystyrene microspheres (5 μm, white, 50 mg / mL), wash twice with MEST buffer (pH = 6.0 containing 0.05% Tween 20), resuspend with MES (pH = 6.0) buffer, add 30 μL of EDC (10 mg / mL) and 30 μL of NHS (10 mg / mL) to 500 mL, mix well, and activate at room temperature for 15 min. After activation, wash twice with PBS (pH = 7.4) buffer, centrifuge and resuspend with 500 ml of PBS buffer, then add 100 μg of mannose, and oscillate at 37°C for 2-4 h; after the reaction is completed, block with 1% BSA solution (diluted with PBS) for 30 min, wash 4 times with PBST (pH = 7.4 containing 0.05% Tween 20) buffer after blocking, resuspend the polystyrene microsphere-CAP-BSA conjugate with PBS buffer (containing 0.5% BSA) and store at 4°C for use.

[0118] The preparation steps of the magnetic nanoparticle-mannose conjugate complex are consistent with those of the polystyrene microsphere carrier-mannose conjugate complex, except that magnetic separation is used instead of centrifugation.

[0119] (2) Bacterial culture and quantification

[0120] E. coli JM109 was cultured in LB broth at 37°C for 16 hours. Salmonella typhimurium was cultured in lactose broth at 37°C for 16 hours. Staphylococcus aureus was cultured in nutrient broth at 37°C for 16 hours. The cultured bacteria were harvested by centrifugation at 6000 rpm for 5 minutes. The supernatant was carefully removed, and the bacterial pellet was left at the bottom, which was then resuspended in PBS. The centrifugation and washing process was repeated twice to ensure complete removal of the culture medium. The number of viable cells was determined by colony counting on agar plates. Before use, the cultured sample was diluted to the desired concentration (104CFU / mL) using PBS.

[0121] (3) Study of the interaction between mannose and bacteria

[0122] 5 μL of mannose-modified polystyrene microspheres and 3 μL of mannose-modified magnetic nanoparticles were mixed with 17 μL of bacterial samples of different concentrations, and shaken at room temperature for 40 minutes. After the reaction was completed, 10 μL of the reaction solution was loaded into the chip at a flow rate of 1.5 μL / min for analysis.

[0123] The microtube weight distribution, image processing, and machine vision recognition steps were consistent with those of Example 1. The results of the interaction between mannose and bacteria are shown in Figure 11 , which show that the method has the ability to study the interaction.

[0124] Example 4: Quantitative detection of novel coronavirus N protein using an on-site mode

[0125] The reaction reagent preparation steps were consistent with those of Example 1, except that the biological recognition molecules used were N protein capture antibodies and detection antibodies.

[0126] (1) Homogeneous clean-up-free double-antibody sandwich immune reaction

[0127] 5 μL of polystyrene microspheres modified with N protein capture antibodies (2.5 mg / mL) and 3 μL of magnetic nanoparticles modified with N protein detection antibodies (500 μg / mL) were mixed with 17 μL of samples. After incubation at room temperature for 10 minutes, 5 μL of the reaction mixture was aliquoted into the sample inlet of the chip, and pumped for analysis using capillary action, as shown in Figure 12 .

[0128] (2) Imaging device photographing

[0129] The image was captured by a mobile phone equipped with a macro lens for signal analysis.

[0130] The microtube weight distribution is substantially consistent with that of step of embodiment 1, except that the number of microtubes in the microtube array is 22.

[0131] (3) Image processing and machine vision recognition

[0132] The process of converting the computer-trained recognition model into TensorFlow Lite for deployment on Android devices involves multiple strategic transformations aimed at achieving cross-platform compatibility and optimizing model performance on mobile devices. The first step in this conversion process is to convert the PyTorch model into the ONNX (Open Neural Network Exchange) format. By converting to ONNX, the model gains flexibility in transferring between various frameworks. Subsequently, with the help of the ONNX-TensorFlow conversion tool, the ONNX model is converted into a TensorFlow model. Finally, the TensorFlow model is converted into the TensorFlow Lite format. The TFLite model is optimized for size and performance, with faster inference time and smaller model size compared to the TensorFlow model. This optimization is crucial for ensuring that the model can run efficiently on Android devices.

[0133] (4) Quantitative detection of N protein

[0134] After the image processing procedure, the recognition model of YOLOv7 deployed on the mobile phone identifies the signal intensity of each microtube based on the pre-calculated weights. With the weighted pixel intensity as the vertical coordinate and the logarithmic value of the N protein concentration as the horizontal coordinate, the relationship between the two can be used for quantitative detection of N protein. In this embodiment, the key parameter concentrations of the biochemical reactions involved are optimized under the best conditions. As shown in the N protein detection results under the optimized conditions Figure 13 , the results show that the minimum detection limit of the method of the present application is 14.3 pg / mL, which is significantly higher than the sensitivity of conventional test strips.

[0135] Some of the steps in the embodiments of the present application can be implemented using software, and the corresponding software programs can be stored in a readable storage medium, such as an optical disc or a hard disk, etc.

[0136] The above description is only the preferred embodiments of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A two-dimensional magnetophoresis separation and detection method, characterized in that: The method comprises: Step 1: Coupling the biorecognition molecule A of the target to be detected on the surface of a non-magnetic polymer insulating microsphere carrier with a certain particle size and / or color; Step 2: coupling the biorecognition molecule B of the target to be detected on the surface of the magnetic nanoparticles, and the biorecognition molecule A matches the biorecognition molecule B; Step 3: Mixing a non-magnetic polymer microsphere carrier of a certain particle size and / or color coupled to the biorecognition molecule A, a magnetic nanoparticle coupled to the biorecognition molecule B, and a test solution containing the target to be detected to perform a homogeneous reaction to obtain a mixed solution containing reaction complexes of different magnetic susceptibilities; Step 4: injecting the mixed solution obtained in step 4 into a microfluidic chip, and performing magnetophoresis separation on the reaction complexes with different magnetic susceptibilities under the action of a magnetic field in a working area of ​​the microfluidic chip, wherein the microfluidic chip includes linearly arranged microtubes for anchoring the reaction complexes with different magnetic susceptibilities; Step 5: Acquire an image of the working area of ​​the microfluidic chip, wherein the image covers each microtube after magnetophoresis separation; Step 6: Based on the position of each microtube, finite element analysis software is used to simulate the physical field of the microfluidic chip, obtain the force value of the reaction complex in each microtube, and convert the force value into the weight of each microtube; Step 7: Use the machine vision recognition model to segment the accumulation area of ​​the microsphere probes in each microtube and calculate the pixel intensity; Step 8: Calculate the weighted pixel intensity of each microtubule based on the weight obtained in step 6 and the pixel intensity obtained by identification in step 7. The weighted pixel intensity is calculated as follows: weight × identified pixel intensity, and the weighted pixel intensities of all microtubules are added together to obtain the total pixel intensity. Step 9: Based on the total pixel intensity, combined with a fitting curve between the total pixel intensity and the concentration of the target to be measured, quantitative detection of the concentration of the target to be measured is achieved.

2. The method according to claim 1, characterized in that The step 6 analyzes the Stokes drag force provided by the fluid and the magnetic force provided by the magnetic field on the reaction complex.

3. The method according to claim 1, characterized in that The non-magnetic polymer insulating microsphere carrier includes: polystyrene microspheres, silica microspheres, polybutadiene microspheres, and polyisoprene microspheres.

4. The method according to claim 1, wherein The particle size of the magnetic nanoparticles is 30-1000 nm.

5. The method according to claim 1, wherein The particle size of the non-magnetic polymer microsphere carrier is 1 to 50 μm.

6. The method according to claim 1, characterized in that The target objects to be detected include: viruses, antibodies, biomarkers, antibiotic molecules, pesticide molecules, veterinary drug molecules, biotoxins, bacteria and various matrix components.

7. The method according to claim 1, characterized in that The biorecognition molecules A and B include: capture antibodies and detection antibodies, detection antibodies and capture antibodies, antibodies and antigens, antigens and antibodies, DNA capture probes and DNA detection probes, DNA detection probes and DNA capture probes.

8. The method according to claim 1, characterized in that The machine vision recognition model is a YOLOv7 image segmentation model.

9. A two-dimensional magnetophoresis separation and detection system, characterized in that: For implementing the method according to any one of claims 1 to 8, the system comprises: a microfluidic chip, an optical imaging device, and a computing device; The microfluidic chip comprises: a stabilization zone (W1), a correction zone (W2), and a working zone (W3) which are sequentially connected and conductive, wherein the correction zone (W2) and the working zone (W3) are located in a magnetic field influence zone (W4), and reaction complexes with different magnetic susceptibilities in a mixed solution are anchored by a linearly arranged microtube array in the working zone (W3) under the action of the magnetic field; The optical imaging device is used to collect images of the working area (W3); The computing device comprises: an image acquisition module, configured to acquire the microtube array image captured by the optical imaging device; a weight calculation module configured to simulate the physical field of the microfluidic chip using finite element analysis software based on the position of each microtube, obtain the force value of the reaction complex in each microtube, and convert the force value into the weight of each microtube; an image segmentation module configured to segment the accumulation area of ​​the microsphere probes in each microtube using a machine vision recognition model and calculate pixel intensity; a pixel intensity calculation module configured to calculate a weighted pixel intensity of each microtubule based on the weight and the identified pixel intensity, wherein the weighted pixel intensity is calculated as follows: weight × identified pixel intensity, and the weighted pixel intensities of all microtubules are summed to obtain a total pixel intensity; The quantitative detection module is configured to achieve quantitative detection of the concentration of the target to be detected based on the total pixel intensity and a fitting curve between the total pixel intensity and the concentration of the target to be detected.

10. The system according to claim 9, characterized in that The optical imaging device includes: a bright field microscope, a dark field microscope, a smart phone, and a portable lens.

Images