Construction method of pathogenic mechanism model of fish nocardia disease

Abstract

The invention discloses a construction method of a fish nocardia disease pathogenic mechanism model, and belongs to the field of biotechnology and disease models. The invention aims to solve the technical problem that a deep causal mechanism is difficult to reveal due to the fact that microscopic dynamic interaction of pathogens and hosts cannot be observed in real time and in situ when an existing animal model and an in-vitro cell model are used for researching the disease. The method comprises the following steps: constructing a fish immune organ in an intelligent micro-fluidic chip integrated with a micro-actuator as a high-bionic-degree physical experiment platform; the living cell imaging system is used for collecting four-dimensional image data of the organoids in the nocardia infection process in real time. According to the method, through normal form transformation from passive observation to active intervention, the microscopic process of the disease can be reproduced in a high-fidelity manner and is regulated and controlled prospectively, so that a causal chain between key pathological events is deeply revealed, and a powerful intelligent research platform is provided for developing a new disease prevention and treatment strategy.

Description

Technical Field

[0001] This invention relates to the field of biomedical research and disease modeling technology, specifically to a method for constructing a pathogenic mechanism model of nocardiosis in fish. Background Technology

[0002] Nocardiosis in fish, primarily caused by *Nocardia seriolae*, is a chronic, debilitating bacterial disease that seriously threatens the global aquaculture industry. A typical pathological feature of this disease is the formation of numerous nodules, or granulomas, within the major immune organs of fish, such as the head kidneys and spleen. These granulomatous structures are the result of a long-term, complex interaction between the host's immune system and the pathogen; their formation, development, and eventual evolution directly determine the course and outcome of the disease. Therefore, a deep understanding of the dynamic spatiotemporal interplay between the pathogen and host immune cells (such as macrophages and lymphocytes) at the microscopic scale during granuloma formation is crucial for revealing the pathogenic mechanism of this disease and finding effective prevention and control strategies.

[0003] Currently, research on the pathogenesis of nocardiosis in fish mainly relies on two traditional technical approaches: whole-animal infection models and in vitro cell culture models. Whole-animal infection models, through direct artificial challenge to fish, can reproduce the macroscopic pathological processes at the physiological system level, exhibiting the highest physiological relevance. However, this model has inherent limitations in revealing microscopic mechanisms. The immune organ environment inside a living animal is complex and opaque, preventing researchers from conducting real-time, high-resolution, continuous in vivo observations of the recruitment, aggregation, and interaction of immune cells after pathogen invasion, as well as the early formation of granulomas, at the cellular and molecular levels. The entire process is like a "black box" for researchers, who can only obtain discrete, static snapshots of information through dissection and sampling at different time points, making it difficult to capture the dynamic causal chain of key pathological events.

[0004] To overcome the limitations of in vivo models in terms of observability, researchers have developed in vitro two-dimensional (2D) cell culture models, such as co-culturing fish macrophages with Nocardia in petri dishes. This model greatly simplifies the research system, making it possible to directly observe the phagocytic behavior of individual cells using a microscope. While this method has value in studying single interactions between specific cells and pathogens, its simulation of real-world in vivo conditions is severely distorted. Traditional 2D culture environments completely ignore the unique three-dimensional spatial structure of in vivo tissues, the physical support of the extracellular matrix, and the collaborative communication between various immune cells. More importantly, the static culture medium environment cannot simulate the dynamic perfusion of tissue fluid in vivo, resulting in drastically different cell metabolism and signal transduction patterns compared to in vivo. These factors collectively lead to conclusions drawn from 2D models that are often difficult to accurately extrapolate to complex in vivo systems.

[0005] In summary, existing technologies face a dilemma when studying the pathogenesis of chronic and complex diseases like nocardiosis in fish: while live animal models offer high physiological relevance, they lack the observability and interventional capability of microscopic processes; and while traditional in vitro cell models facilitate observation, they severely deviate from the real in vivo microenvironment, significantly diminishing the biological significance of the research results. Therefore, there is an urgent need in this field for a novel technological solution that can reproduce the key microscopic processes of disease development with high spatiotemporal resolution in a highly biomimetic three-dimensional microenvironment, and that can shift from passive observation to active, intelligent intervention, thereby enabling the precise inference of the core causal mechanisms determining the course of the disease from a complex array of correlational phenomena. Summary of the Invention

[0006] The technical problem this invention aims to solve is that existing methods for studying the pathogenesis of nocardiosis in fish have limitations. Traditional animal infection models cannot observe the dynamic interaction between pathogens and hosts at the microscopic level in real time and in situ, resembling a "black box." Furthermore, in vitro two-dimensional cell culture models severely simplify the three-dimensional tissue structure and complex cell communication microenvironment in vivo, resulting in insufficient physiological relevance. These methods struggle to reveal the dynamic causal chain from molecular perturbations to histopathological evolution.

[0007] To address the aforementioned technical problems, this invention provides a novel method for constructing a model of the pathogenesis of nocardiosis in fish that enables dynamic simulation, prediction, and active intervention.

[0008] This invention provides a method for constructing a pathogenic mechanism model of nocardiosis in fish. This method integrates fish immune organoids, intelligent microfluidic chips, and predictive digital twin technology to construct a closed-loop intelligent research system encompassing "perception-cognition-execution." The method includes the following steps: First, a microfluidic chip is provided, and fish immune organoids are constructed within the microfluidic chip. In a preferred embodiment, the microfluidic chip integrates microactuators for performing subsequent microenvironment interventions, and internally designs an organoid culture chamber for three-dimensional culture, a nutrient channel for continuous nutrient supply, and an intervention channel for precise substance delivery. Further, the fish immune organoid is a three-dimensional cell complex containing macrophages, lymphocytes, and stromal cells, formed by self-organizing primary immune cells from the head kidney or spleen of fish in a three-dimensional hydrogel matrix. This structure can better simulate the in vivo immune microenvironment.

[0009] Next, the fish immune organoids were infected with Nocardia, and dynamic image data of the fish immune organoids during the infection process were acquired in real time and continuously using a live-cell imaging system. These data contained rich information about the pathogen and host cells in three-dimensional space and time.

[0010] Then, based on the obtained dynamic image data, a digital twin model synchronized in real time with the fish immune organoids in the physical world is constructed, and the digital twin model is used to predict the system state of the fish immune organoids at future time points. Specifically, this step includes transforming the dynamic image data into a model that can comprehensively describe the system at any given time using image processing and feature extraction algorithms. The mathematical representation of a state, i.e., a state vector. The state vector It contains state information of all identified host cells and pathogens within the fish's immune organoids. To more accurately describe the system state, in one specific implementation, the state vector... The individual host cell state contained Described by a vector, which specifically contains the three-dimensional spatial position vector of the cell. Instantaneous velocity vector Morphological parameters used to describe its size and shape and biological state parameters reflecting its physiological activity. Similarly, the state vector Individual pathogen states included It is also described by a vector, which specifically contains the three-dimensional spatial location vector of the pathogen. Instantaneous velocity vector morphological parameters and biological state parameters .

[0011] After constructing the digital twin model, this invention utilizes a dynamic prediction model. For the current moment state vector Perform calculations to obtain a short time step in the future. Predicting the state of the system afterwards The relationship in this prediction process can be defined by the following formula: ; in, For the dynamic prediction model The set of internal parameters obtained through learning or training.

[0012] Finally, based on the predicted system state, an intervention command is generated, driving the microactuator within the microfluidic chip to perform microenvironmental intervention on the fish's immune organoids. In one specific embodiment, the step of generating the intervention command includes: first, assessing the probability of a set of preset key pathological events based on the predicted system state; then, when the probability of a key pathological event exceeds a preset threshold, a targeted intervention command is generated. To achieve precise control, the intervention command may include information such as the intervention type (e.g., what substance to release), the spatial coordinates of the intervention target (its specific location within the chip), and intervention parameters (e.g., release concentration and duration). The judgment process can be limited by the following conditions: ; Where Prob is the probability calculation function. For the first Key pathological events (such as granuloma formation), The target spatial coordinates where the event occurred. For the predicted system state, In response to the incident A preset probability threshold. When this condition is met, the system will automatically trigger the corresponding physical intervention.

[0013] This invention establishes an intelligent research paradigm that moves from passive observation to proactive intervention through the aforementioned methods. It not only reproduces the Nocardia infection process with high fidelity, but more importantly, through a closed-loop "prediction-intervention" regulation, it can proactively explore and verify the causal relationships between different biological events, thereby profoundly revealing the intrinsic mechanisms of disease development and providing an unprecedentedly powerful tool for developing new prevention and control strategies.

[0014] This invention provides a method for constructing a pathogenic mechanism model of nocardiosis in fish. It has the following beneficial effects: 1. This invention significantly improves the biofidelity of in vitro models by employing an intelligent microfluidic chip integrating dynamic perfusion channels and constructing fish immune organoids within it. This technical solution not only reproduces the three-dimensional spatial configuration and cellular diversity of in vivo tissues but also provides cells with a dynamic microenvironment that more closely resembles their physiological state by simulating the continuous flow of tissue fluid. This overcomes the fundamental deficiency of traditional 2D cell culture models, which suffer from insufficient reliability of experimental results due to environmental distortion, enabling the cell behavior and disease evolution processes in the model to more realistically reflect the conditions in vivo.

[0015] 2. This invention directly couples an optically transparent microfluidic chip with a live-cell imaging system, achieving for the first time continuous, non-destructive four-dimensional observation of the microscopic processes of nocardiosis in fish at high spatiotemporal resolution. This design completely opens the long-standing "black box" in traditional live animal model research, allowing researchers to intuitively track the complete dynamic trajectory from single pathogen invasion and immune cell response to the formation of early granulomatous structures. This shift from discrete static snapshots to continuous dynamic imaging provides unprecedented capabilities for capturing and understanding fleeting critical pathological events.

[0016] 3. The most valuable innovation of this invention lies in the introduction of a digital twin system, achieving a paradigm shift from passive observation to proactive intervention. This system is not only a digital record of physical experiments, but also possesses the ability to prospectively predict the future state of a disease through a dynamic predictive model. This predictive capability allows the system to provide early warnings and trigger precise physical interventions before key pathological events (such as irreversible tissue fibrosis) actually occur, which is completely impossible with traditional research methods.

[0017] 4. By constructing an intelligent closed loop of "prediction-intervention-response," this invention establishes a highly efficient causal relationship inference mechanism. The system can autonomously conduct numerous "perturbation-response" experiments at the microscopic scale, that is, after predicting a certain trend, it actively applies a micro-perturbation and accurately records the subsequent evolutionary trajectory of the system. By comparing the differences in the system state before and after intervention, the true role of specific cellular behaviors or signaling pathways in the disease process can be directly and powerfully revealed, thereby refining complex correlation phenomena into a clear causal chain.

[0018] 5. The dynamic pathogenic mechanism model ultimately produced by this invention is itself a powerful simulation platform that has been repeatedly verified and calibrated in the physical world. This model transcends static signal pathway diagrams and can be used for low-cost, extremely fast "in-computer" simulations to test the effectiveness of different candidate drug targets or evaluate the potential effects of multiple combination therapy strategies. This significantly shortens the cycle from basic mechanism research to potential clinical application translation, providing a novel and efficient research tool for the development of precision control strategies for nocardiosis in fish. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall system architecture according to an embodiment of the present invention; Figure 2 This is a three-dimensional structural schematic diagram of the intelligent microfluidic chip of the present invention; Figure 3 This is a schematic cross-sectional view of the intelligent microfluidic chip of the present invention; Figure 4 This is a flowchart of the closed-loop intelligent control method of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The embodiments described herein are merely illustrative and are not intended to limit the scope of protection of this invention.

[0021] refer to Figures 1 to 4 This invention provides a method for constructing a pathogenic mechanism model of nocardiosis in fish. This method is implemented through an integrated closed-loop intelligent system, which mainly includes an intelligent microfluidic chip as a physical platform, a live cell imaging system for real-time observation, and a digital twin system as the core of computing and control.

[0022] The method for constructing the pathogenic mechanism model of nocardiosis in fish may include the following steps: First, prepare and provide the intelligent microfluidic chip, and construct fish immune organoids therein; then, initiate the Nocardia infection process in the system, and continuously acquire dynamic image data using the live-cell imaging system; next, the digital twin system receives and processes the data, constructs a digital twin model, and predicts future states; finally, the digital twin system generates and sends intervention commands based on the prediction results, controlling the micro-actuator array within the intelligent microfluidic chip to perform precise microenvironment intervention.

[0023] This embodiment will first describe in detail the hardware of the intelligent closed-loop system and the preparation process of biological samples.

[0024] The process first involves the design and fabrication of the intelligent microfluidic chip. The main body of the intelligent microfluidic chip is made of a non-toxic, light-transmitting material, such as polydimethylsiloxane (PDMS), and is fabricated using standard soft lithography and multilayer bonding processes. The internal structure of the chip is precisely designed to support three-dimensional cell culture and dynamic microenvironment regulation.

[0025] Specifically, the intelligent microfluidic chip contains one or more independent organoid culture chambers. The geometry and size of these chambers are optimized to facilitate the formation and stabilization of immune organoids. A multi-layered microfluidic channel network is designed around each organoid culture chamber. One layer is a nutrient channel, connected to the bottom of the organoid culture chamber via a porous polymer film. This channel is used to continuously perfuse basal culture medium at extremely low flow rates, simulating the osmotic nutrient environment of in vivo tissue fluid, while effectively removing cellular metabolic products.

[0026] The other layer is an intervention channel, which has independent fluid inlets and outlets, with its branch ends precisely opening to the side or top of the organoid culture chamber. This intervention channel is used to precisely and controllably introduce pathogen suspensions, drug solutions, or other bioactive molecules at specific stages of the experiment. To achieve active and precise intervention in the microenvironment, a microactuator array, uniformly controlled by an external controller, is integrated at key branch nodes of the intervention channel. In this embodiment, the microactuator array is a set of piezoelectric microvalves, capable of rapid and precise switching and flow control of nano-scale fluids in specific channel branches based on electrical signal commands from the digital twin system.

[0027] Simultaneously with hardware preparation, biological samples were prepared. This process included the preparation of fish immune organoids. Healthy largemouth bass (Micropterus salmoides) were selected, and their head kidney tissue was obtained under sterile conditions. The tissue was dissociated into a single-cell suspension through mechanical shearing combined with enzymatic digestion using collagenase and dispersant enzymes. Subsequently, density gradient centrifugation was used to remove red blood cells and tissue debris, enriching the primary immune cells containing various types, including macrophages, lymphocytes, and stromal cells.

[0028] The obtained primary immune cells were uniformly mixed with a pre-cooled liquid biocompatible three-dimensional hydrogel matrix (e.g., Matrigel) at a predetermined volume ratio. The cell-hydrogel mixture was immediately and precisely injected into the organoid culture chamber of the intelligent microfluidic chip using a micropipette. The chip was placed in a cell culture incubator, allowing the hydrogel to thermally solidify and stably embed the cells in three-dimensional space. Subsequently, the chip was connected to a microfluidic perfusion system, continuously perfusing an induction medium containing a specific combination of cytokines and growth factors through the nutrient channels. Under the combined action of the three-dimensional structure and the dynamic fluid microenvironment, the cells proliferated, migrated, and communicated, and after a period of culture, self-organized to form fish immune organoids that structurally and functionally mimic in vivo immune tissues.

[0029] In addition, fluorescently labeled Nocardia bacteria for infection experiments need to be prepared. A virulent strain of fish-derived Nocardiaseriolae is inoculated into liquid culture medium for amplification. An expression plasmid carrying the green fluorescent protein (GFP) gene is introduced into the bacteria via electroporation, and recombinant strains capable of stably expressing GFP are obtained through antibiotic selection. Before the experiment, the fluorescently labeled strain is cultured to the logarithmic growth phase and resuspended in sterile buffer to a predetermined concentration for subsequent infection experiments.

[0030] In this embodiment, once the fish immune organoids within the intelligent microfluidic chip have matured and reached a stable state, the dynamic infection model can be initiated and high-content data can be acquired in real time. This process aims to accurately simulate the Nocardia invasion event and continuously record the entire interaction between the pathogen and the host immune system with high spatiotemporal resolution.

[0031] First, the infection model is initiated. A microinjection pump connected to the inlet of the intervention channel on the intelligent microfluidic chip is used to inject a fluorescently labeled Nocardia suspension, prepared in the previous steps and with a precisely known concentration, at a preset, slow, and stable flow rate. The bacterial suspension flows along the intervention channel and is released at its opening into the surrounding microenvironment of the fish's immune organoids, thereby forming a controllable initial infection focus in a localized area, simulating the colonization process of pathogens in tissues.

[0032] Upon initiation of infection, the entire intelligent microfluidic chip system is immediately transferred to the stage of the live-cell imaging system. In this invention, the live-cell imaging system is a laser scanning confocal microscope equipped with a high-sensitivity camera and integrates an environmental control chamber. This control chamber provides the intelligent microfluidic chip with a constant temperature, humidity, and gas atmosphere, ensuring that the cells within the organoid maintain good physiological activity throughout imaging cycles lasting several hours or even days.

[0033] Subsequently, imaging parameters were set and optimized to achieve non-destructive, high-content information acquisition of the dynamic infection process. A suitable objective lens was selected, such as a 20x or 40x water microscope with a long working distance, to balance a sufficiently large field of view with the spatial resolution required to resolve individual cells. Multicolor fluorescence acquisition channels were set up; for example, one channel was used to excite and acquire the green fluorescent protein (GFP) signal carried by the pathogen, and another channel was used to excite and acquire the blue fluorescence signal emitted by a dye pre-labeled on the host cell nucleus (such as DAPI or Hoechst33342).

[0034] To obtain complete three-dimensional spatial information, a Z-stack scan was performed on the region containing the fish immune organoids. The Z-stack scan range needed to completely cover the thickness of the entire organoid, and the scan step size was set according to the axial resolution of the objective lens to ensure Nyquist sampling in the Z-axis direction.

[0035] To capture dynamic processes, the system repeats a complete three-dimensional multicolor fluorescence image scan at fixed time intervals (e.g., every 5 or 10 minutes). This results in a continuous output of a four-dimensional (4D) image dataset, which, in time-series format, records the precise location, morphology, and fluorescence intensity information of all fluorescently labeled pathogens and host cells at every time point (t) in three-dimensional space (x, y, z). This series of raw data streams is transmitted in real-time to the digital twin system, providing the necessary high-quality input for subsequent quantitative analysis, model building, and closed-loop control.

[0036] In this embodiment, the digital twin system is the core of the entire closed-loop intelligent control. Its function is to transform the high-dimensional raw image data stream acquired by the live-cell imaging system into a structured mathematical description, and based on this, to construct a dynamic model that can predict the future, and finally generate feedforward intervention commands to actively control the physical system.

[0037] The first step in this process is to process the real-time acquired four-dimensional image data and vectorize it into a precise representation of the system state. Each frame of three-dimensional image transmitted from the live-cell imaging system first passes through an image preprocessing module, which performs operations including Gaussian filtering for noise reduction and background signal correction to improve the accuracy of subsequent analysis. Subsequently, the preprocessed image is fed into a deep learning-based instance segmentation model, such as a U-Net or Mask R-CNN network pre-trained on a similar dataset. This model can accurately identify and segment each individual host cell nucleus and the outline of each pathogen or pathogen colony in the image.

[0038] After segmentation, the system extracts features and tracks each identified object across time frames. By calculating the geometric center of each segmented object, its three-dimensional spatial position can be determined. By comparing the positional changes of the same object in consecutive time frames, its instantaneous velocity vector can be calculated. Simultaneously, morphological parameters such as volume, surface area, and sphericity, as well as biological state parameters such as the average fluorescence intensity within its fluorescence channels, can be calculated. A multi-object tracking module based on Kalman filtering or the Hungarian algorithm is responsible for establishing a unique identity correspondence for each object across consecutive time frames.

[0039] Through the above processing, the state of the physical system at any time t is completely abstracted into a comprehensive state vector S(t). The mathematical form of this vector is:

[0040] in, and These represent the total number of host cells and pathogens identified and tracked at time t, respectively. Among them, a single host cell... The state is described by a vector:

[0041] in, Let be the three-dimensional spatial position vector of cell i. Let it be its instantaneous velocity vector. Its morphological parameter set, This is its set of biological state parameters. Similarly, a single pathogen or colony... The state is also determined by a vector. The process is described below. The second step is to utilize the real-time generated state vector. To drive a dynamic prediction model The operation of this model constitutes the cognitive core of digital twins. In this invention, the dynamic prediction model... A graph neural network (GNN) architecture is employed. This model treats each cell and pathogen as a node in a graph, with spatial proximity and interactions between nodes defined as edges. Trained on a large amount of historical data, the model learns the intrinsic biological and physical laws governing cell migration, proliferation, interactions, and pathogen growth and spread. The model's function is based on the state vector at the current time t. Calculate and output the time step of a short future time. Predicting the state of the system afterwards This prediction process can be described by the following formula:

[0042] in, The dynamic prediction model is A set of internal parameters that are fixed after training.

[0043] The third step in this process is to transform the model's predictive power into the ability to intervene in the physical world, thus forming a closed loop. A set of key pathological events is predefined within the system. ,For example, This can be defined as "the density of macrophages in a local area exceeds a specific value". This can be defined as "early granulomatous core structure formation," etc. The digital twin system will be based on the predicted system state. Real-time calculation of arbitrary spatial coordinates within organoids Key pathological events occurred at the site. The probability of.

[0044] The system has a built-in control law function. This function continuously monitors the predicted probability of all key pathological events. When a certain event... At a specific location The predicted probability exceeded a preset trigger threshold. At this time, the control law is activated. This activation condition can be limited by the following inequality:

[0045] in, This represents a probability calculation function. Once this condition is met, the digital twin system will immediately generate a specific intervention command. .

[0046] The intervention instructions It is a structured data packet that contains all the information needed to perform the intervention, such as the type of intervention (e.g., releasing immunosuppressants) and the target spatial coordinates of the intervention. (corresponding to one or more specific units in the microactuator array), and intervention parameters (e.g., the concentration and duration of the released solution). This instruction is sent via a communication interface to the external controller of the fully-fledged intelligent microfluidic chip, which interprets it as a specific electrical signal to drive the corresponding unit in the microactuator array to perform precise physical operations, thereby enabling proactive and precise regulation of the target microenvironment before key pathological events actually occur.

[0047] In this embodiment, through the continuous operation of the aforementioned steps, the entire closed-loop intelligent system will generate massive amounts of data rich in biological significance. After the experiment, the final step is to perform in-depth integration and analysis of this data, and ultimately output a dynamic pathogenic mechanism model with causal inference capabilities.

[0048] During the experiment, the digital twin system not only recorded the natural evolution trajectory of the physical system (i.e., the control group experiment under no intervention or baseline conditions), but also fully recorded the predicted state of the system before each closed-loop intervention was triggered, the generated intervention instructions, and the actual response and subsequent evolution path of the physical system after the intervention. This design enables the system to conduct multiple sets of "virtual experiments" and "physical experiments" in parallel, and directly establish the correlation between specific interventions and system evolution results.

[0049] The first step in data analysis is data integration and causal relationship inference. All collected four-dimensional image data, processed state vector time series, and intervention event logs are aligned and integrated. By comparing and analyzing the differences in system evolution trajectories under different intervention conditions, the impact of specific interventions (e.g., the premature release of a specific cytokine inhibitor in a predicted cell aggregation hotspot) on the probability and progression of subsequent pathological events (e.g., the formation of mature granulomas) can be quantitatively assessed. This perturbation-response-based analysis method can effectively identify key regulatory pathways with strong causal orientation from complex correlation networks.

[0050] The second step in data analysis is to construct and visualize a dynamic model of the pathogenic mechanism. All significant causal relationships inferred in the previous step are summarized to construct a directed acyclic graph or dynamic Bayesian network. In this network model, nodes represent different biological entities (such as specific types of cells or pathogen colonies) or biological events (such as cell activation, chemotactic migration, or granuloma formation), and directed edges represent the causal driving relationships between them. The weights of the edges can be defined by the quantitative impact obtained in the preceding analysis.

[0051] Ultimately, the pathogenic mechanism model constructed in this invention is not a static pathway diagram, but a dynamic and computable model. It can be visualized as an interactive four-dimensional animation, intuitively demonstrating how various cell populations interact, how key signals are transmitted spatiotemporally, and ultimately lead to a specific pathological outcome under specific initial conditions, starting from Nocardia invasion.

[0052] Furthermore, the model itself can also serve as a powerful simulation platform. Researchers can perform purely "in-computer" simulations within this model, which has already been calibrated with physical experimental data. For example, they can simulate the inhibitory effect of a new candidate drug molecule on a key node in the system and predict the long-term impact of this intervention on the entire disease process. This provides an efficient and low-cost means for rapidly screening drug targets and optimizing treatment strategies, thereby achieving the ultimate goal of this invention: to construct a pathogenic mechanism model that can deeply analyze and guide disease intervention.

Claims

1. A method for constructing a pathogenic mechanism model of nocardiosis in fish, characterized in that, Includes the following steps: A microfluidic chip is provided, and fish immune organoids are constructed within the microfluidic chip; The fish immune organoids were infected with Nocardia, and dynamic image data of the fish immune organoids were acquired in real time. Based on the dynamic image data, a digital twin model synchronized with the fish immune organoid is constructed, and the digital twin model is used to predict the system state of the fish immune organoid at future time points. Based on the predicted system state, an intervention command is generated, and the microactuator in the microfluidic chip is driven to perform microenvironmental intervention on the fish immune organoids.

2. The method according to claim 1, characterized in that, The step of providing a microfluidic chip includes: providing a microfluidic chip that integrates microactuators and includes an organoid culture chamber, a nutrient channel and an intervention channel.

3. The method according to claim 1, characterized in that, The step of acquiring dynamic image data of the fish immune organoids in real time and constructing a digital twin model based on the dynamic image data includes: processing the dynamic image data into a state vector S(t) describing the state of the system at time t, wherein the state vector S(t) includes the state information of the host cells and pathogens within the fish immune organoids.

4. The method according to claim 1 or 3, characterized in that, The step of using the digital twin model to predict the system state of the fish immune organoids at future time points includes: using a dynamic prediction model F to calculate the state vector S(t) at the current time t to obtain the state at a future time step. Predicting the state of the system afterwards Their relationship satisfies: ; Wherein, θ is the set of parameters of the dynamic prediction model F.

5. The method according to claim 1, characterized in that, The step of generating intervention instructions based on the predicted system state includes: Based on the predicted system state, assess the probability of the occurrence of preset key pathological events; When the probability of the occurrence of the key pathological event exceeds a preset threshold, the intervention instruction is generated.

6. The method according to claim 5, characterized in that, The step of assessing the probability of the occurrence of a preset key pathological event and generating an intervention instruction when the probability exceeds a preset threshold is determined by the following conditions: ; in, This is a probability calculation function. For the first Key pathological events, The target spatial coordinates where the event occurred. For the predicted system state, In response to the incident Preset probability threshold.

7. The method according to claim 1 or 5, characterized in that, The intervention instruction includes the following information: intervention type, spatial coordinates of the intervention target, and intervention parameters.

8. The method according to claim 3, characterized in that, The state vector Described by a vector, which contains the three-dimensional spatial position vector of the cell. Instantaneous velocity vector morphological parameters and biological state parameters .

9. The method according to claim 3 or 8, characterized in that, The state vector Pathogen states included Described by a vector, which contains the three-dimensional spatial location vector of the pathogen. Instantaneous velocity vector morphological parameters and biological state parameters .

10. The method according to claim 1, characterized in that, The fish immune organoids are three-dimensional cell complexes containing macrophages, lymphocytes, and stromal cells, which are formed by the self-organization of primary immune cells from the fish head kidney or spleen.

Images