Dynamic coprophilous fungus transplantation donor and acceptor matching method, device, equipment and medium

Abstract

The embodiment of the invention discloses a coprophilous fungus transplantation donor and acceptor dynamic matching method, device and equipment and a medium, and the method comprises the steps: responding to a donor and acceptor matching request, and obtaining the monitoring data of a to-be-matched acceptor after coprophilous fungus transplantation at the current preset sampling time; wherein the monitoring data after coprophilous fungus transplantation is determined on the basis of the coprophilous fungus transplantation operation of a to-be-matched receptor completed by the coprophilous fungus flora of any target donor, and the target donor is determined by performing adaptability matching on the gene feature information, the immune feature information and the clinical feature information of the to-be-matched receptor and a plurality of to-be-selected donors; and inputting monitoring data after coprophilous fungi transplantation into the transplantation strategy control model, outputting at least one of a donor switching strategy and a dose adjustment strategy for adjusting a to-be-matched receptor in a next time period, realizing comprehensive compatibility evaluation through multi-modal data integration, continuously optimizing a treatment scheme by means of a dynamic feedback mechanism, and improving the compatibility of the to-be-matched receptor. And the accuracy and adaptability of coprophilous fungus transplantation treatment are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intestinal flora transplantation, and in particular to a fecal microbiota transplantation donor-recipient dynamic matching method, device, equipment and medium. BACKGROUND

[0002] Fecal microbiota transplantation (FMT) is a treatment method for reconstructing the intestinal micro-ecosystem of the recipient by transplanting the intestinal flora of a healthy donor, which has shown significant efficacy in diseases such as recurrent Clostridium difficile infection and inflammatory bowel disease. Its core mechanism is to regulate the immune metabolism function of the host through the implantation of microbial communities and restore the balance of the intestinal micro-ecosystem. With the development of microbiome, immunology and artificial intelligence technology, FMT is gradually evolving from empirical treatment to precision and individualization, but multi-dimensional data integration and dynamic optimization are still key difficulties for technological breakthrough.

[0003] The current FMT matching method mainly relies on a static matching strategy based on microbiome similarity analysis, that is, a donor is determined once before fecal microbiota transplantation through a similarity evaluation algorithm, so that the intestinal flora of the donor provides a healthy flora for the recipient.

[0004] However, the above-mentioned FMT matching method has the following problems: on the one hand, the above-mentioned FMT matching method relies too much on microbiome data, and has the problem of single data dimension, which makes it difficult to comprehensively evaluate the biological compatibility of the donor and the recipient; on the other hand, due to the static matching strategy, it cannot respond to the dynamic changes of the recipient's flora after transplantation and the evolution of the clinical state, resulting in treatment lag. Based on this, the existing FMT matching method has the dual limitations of the static matching strategy being unable to adapt to the dynamic changes after transplantation and the single data dimension leading to incomplete evaluation of biological compatibility. SUMMARY

[0005] The present application provides a fecal microbiota transplantation donor-recipient dynamic matching method, device, equipment and medium to achieve comprehensive compatibility evaluation through multi-modal data integration and continuously optimize the treatment plan with the help of a dynamic feedback mechanism, significantly improving the precision and adaptability of fecal microbiota transplantation treatment.

[0006] In a first aspect, the present application embodiment provides a fecal microbiota transplantation donor-recipient dynamic matching method, which comprises:

[0007] In response to the donor-recipient matching request, obtain post-fecal bacteria transplantation monitoring data of the recipient to be matched at a current preset sampling time; wherein the post-fecal bacteria transplantation monitoring data includes recipient symptom score and flora colonization rate, and the post-fecal bacteria transplantation monitoring data is determined based on the fecal flora of any one target donor after the fecal bacteria transplantation surgery is completed on the recipient to be matched, and the target donor is determined based on the genetic feature information, immune feature information and clinical feature information of the recipient to be matched and a plurality of candidate donors.

[0008] Input the post-fecal bacteria transplantation monitoring data into a transplantation strategy control model, and output a transplantation adjustment strategy for adjusting the recipient to be matched in the next period; wherein the transplantation adjustment strategy includes at least one of a donor switching strategy and a dose adjustment strategy.

[0009] In a second aspect, the embodiments of the present application also provide a fecal bacteria transplantation donor-recipient dynamic matching device, which comprises:

[0010] A monitoring data acquisition module is configured to obtain post-fecal bacteria transplantation monitoring data of the recipient to be matched at a current preset sampling time in response to a donor-recipient matching request; wherein the post-fecal bacteria transplantation monitoring data includes recipient symptom score and flora colonization rate, and the post-fecal bacteria transplantation monitoring data is determined based on the fecal flora of any one target donor after the fecal bacteria transplantation surgery is completed on the recipient to be matched, and the target donor is determined based on the genetic feature information, immune feature information and clinical feature information of the recipient to be matched and a plurality of candidate donors.

[0011] An adjustment strategy determination module is configured to input the post-fecal bacteria transplantation monitoring data into a transplantation strategy control model, and output a transplantation adjustment strategy for adjusting the recipient to be matched in the next period; wherein the transplantation adjustment strategy includes at least one of a donor switching strategy and a dose adjustment strategy.

[0012] In a third aspect, the embodiments of the present application also provide an electronic device, which comprises:

[0013] One or more processors;

[0014] A storage device is configured to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the fecal bacteria transplantation donor-recipient dynamic matching method according to any one of the embodiments of the present application.

[0015] In a fourth aspect, the embodiments of the present application also provide a storage medium containing computer executable instructions, which are used to execute the fecal bacteria transplantation donor-recipient dynamic matching method according to any one of the embodiments of the present application when executed by a computer processor.

[0016] The technical scheme of the embodiment of the present application, in response to the donor-recipient matching request, acquires the fecal flora transplantation post-monitoring data of the to-be-matched recipient at the current preset sampling time, wherein the fecal flora transplantation post-monitoring data includes the recipient symptom score and the flora colonization rate, the fecal flora transplantation post-monitoring data is determined after the to-be-matched recipient completes the fecal flora transplantation surgery based on the fecal flora of any one target donor, and the target donor is determined by adaptively matching the genetic feature information, the immune feature information and the clinical feature information of the to-be-matched recipient and a plurality of candidate donors; further, the fecal flora transplantation post-monitoring data is input into the transplantation strategy control model, and a transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period is output; wherein the transplantation adjustment strategy includes at least one of the donor switching strategy and the dose adjustment strategy. The technical scheme provided in this embodiment is based on the dynamic feedback mechanism of the post-transplantation real-time monitoring data, continuously optimizes the donor selection and dose adjustment through the transplantation strategy control model, solves the pain point that the static matching cannot adapt to the dynamic changes after transplantation, in addition, integrates the multi-modal data such as genetic features, immune features and clinical features, breaks through the limitation of single microbiome data; through the multi-modal data integration, comprehensive compatibility evaluation is realized, and with the help of the dynamic feedback mechanism, the treatment scheme is continuously optimized, which significantly improves the precision and adaptability of FMT treatment. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical scheme of the exemplary embodiments of the present application, the drawings needed in the description of the embodiments will be briefly introduced. Obviously, the drawings introduced are only a part of the drawings of the embodiments to be described by the present application, and not all the drawings. Those skilled in the art can obtain other drawings from these drawings without creating labor.

[0018] Figure 1 The flowchart of the fecal flora transplantation donor-recipient dynamic matching method provided by the embodiment of the present application is shown in the figure.

[0019] Figure 2 The flowchart of another fecal flora transplantation donor-recipient dynamic matching method provided by the embodiment of the present application is shown in the figure.

[0020] Figure 3 The flowchart of another fecal flora transplantation donor-recipient dynamic matching method provided by the embodiment of the present application is shown in the figure.

[0021] Figure 4 The structural diagram of the fecal flora transplantation donor-recipient dynamic matching device provided by the embodiment of the present application is shown in the figure.

[0022] Figure 5 The structural diagram of the electronic device provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0023] The application will be described in further detail below with reference to the drawings and embodiments. It is to be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application. In addition, it should be noted that only the parts related to the present application are shown in the drawings for ease of description.

[0024] Before introducing the embodiments of the present application, the application scenarios in which the embodiments of the present application are implemented can be exemplarily described. The embodiments of the present application can be applied to any situation in which it is necessary to dynamically monitor after fecal bacteria transplantation and to real-time adjust the donor or transplantation strategy according to the symptoms and flora colonization of the recipient. Among them, fecal bacteria transplantation is a medical method of extracting and processing functional flora in the feces of a healthy donor through a specific technology and transplanting them into the intestinal tract of a patient to reconstruct the intestinal microecological balance. The core purpose is to inhibit pathogenic bacteria and repair intestinal flora disorder by introducing diversified beneficial microorganisms, so as to treat diseases related to intestinal microbial imbalance. The key steps of the technology include donor screening, feces processing, and delivering flora to the intestinal tract of a patient through oral capsules, nasoenteric tubes, colonoscopes, etc. The curative effect depends on the functional integrity of the donor flora and the compatibility of the recipient intestinal environment. The present embodiment aims to dynamically monitor the symptom score and flora colonization rate of the recipient after fecal bacteria transplantation, combine genetic, immune, and clinical characteristic information, and use a transplantation strategy control model to real-time optimize donor selection and transplantation scheme, so as to improve the precision and curative effect stability of fecal bacteria transplantation, reduce transplantation failure or adverse reactions caused by donor-recipient mismatch, and realize individualized and adaptive fecal bacteria transplantation treatment management.

[0025] Figure 1 A flowchart of a fecal bacteria transplantation donor-recipient dynamic matching method provided by the embodiments of the present application is shown. The present embodiment can be applied to any situation in which it is necessary to dynamically monitor after fecal bacteria transplantation and to real-time adjust the donor or transplantation strategy according to the symptoms and flora colonization of the recipient. The method can be executed by a fecal bacteria transplantation donor-recipient dynamic matching device. The device can be realized in the form of software and / or hardware. The hardware can be an electronic device, such as a mobile terminal, a PC terminal, or a server, etc.

[0026] As shown in FIG. 1, the fecal bacteria transplantation donor-recipient dynamic matching method includes the following steps. Figure 1

[0027] S110, in response to a donor-recipient matching request, obtaining fecal bacteria transplantation post-monitoring data of a to-be-matched recipient at a current preset sampling time.

[0028] ​The donor-recipient matching request refers to an instruction or requirement initiated by a medical system or related party for dynamic matching of a specific recipient (patient) and potential donor (fecal flora provider) during fecal flora transplantation. The recipient to be matched refers to a patient individual who needs to be evaluated for donor matching during fecal flora transplantation treatment. The recipient has received fecal flora transplantation based on initial matching, but still needs to further optimize donor selection or transplantation strategy according to postoperative dynamic monitoring data. The current preset sampling time refers to a pre-set key monitoring time node for evaluating the postoperative state of the recipient. The time point is determined based on clinical treatment specifications or individualized schemes, and is designed to systematically collect the post-transplantation monitoring data of the recipient at that time, to provide timely input data for the dynamic matching model, to support the stage-by-stage evaluation of the transplantation effect and subsequent strategy adjustment decisions.

[0029] The post-fecal flora transplantation monitoring data refers to a set of quantitative indicators collected at a pre-set time point after fecal flora transplantation for evaluating the treatment effect of the recipient. The post-fecal flora transplantation monitoring data includes the recipient symptom score and the flora colonization rate. The recipient symptom score refers to an index parameter that quantitatively characterizes the improvement degree of the clinical symptoms of the fecal flora transplantation recipient through a standardized evaluation system. The scoring system is based on pre-defined clinical evaluation standards, and through the collection of physiological indicators, subjective symptom reports and other data of the recipient, a specific algorithm is used to convert the treatment effect into a quantifiable numerical value. The flora colonization rate refers to a proportion index of the successful colonization of the donor-derived flora in the recipient's intestinal tract determined by microbiome analysis technology. This parameter quantitatively represents the implantation efficiency and stability of the transplanted flora in the recipient's intestinal microecosystem. As a core biological indicator for evaluating the effect of fecal flora transplantation, it reflects the survival status and ecological niche occupation degree of the donor functional flora in the recipient's intestinal tract, and the numerical change is directly related to the microbiological efficacy of transplantation treatment, providing objective microbial colonization evidence for clinical decision-making.

[0030] More specifically, the post-fecal bacteria transplantation monitoring data is determined based on the fecal flora of any one target donor after the fecal bacteria transplantation surgery is completed on the matched recipient, and the target donor is determined based on the genetic feature information, immune feature information and clinical feature information of the matched recipient and a plurality of candidate donors. The target donor refers to the fecal flora provider selected from the candidate donor library through multi-dimensional feature matching and most compatible with the matched recipient. The selection is determined based on the multi-parameter comprehensive evaluation results of the genetic features, immune features and clinical features of the donor and recipient, and the core feature is the optimal biological compatibility of the recipient in terms of microbial community composition, function and host interaction. The candidate donor refers to the set of potential fecal flora providers that enter the final matching evaluation stage after preliminary screening. The genetic feature information refers to the set of genetic parameters used to evaluate the biological compatibility of the two parties, mainly including the genetic matching degree data of the donor and the recipient in terms of specific functional gene sites, HLA genotypes and microbial group feature genes. The immune feature information refers to the set of key immunological parameters used to evaluate the immune compatibility of the donor and the recipient, mainly including the feature data of the donor and the recipient in terms of immune cell composition, cytokine spectrum, immune response mode and mucosal immune state. The clinical feature information refers to the set of medical parameters used to evaluate the clinical compatibility of the donor and the recipient in the fecal bacteria transplantation matching process, mainly including disease type, disease stage, medication history, complications and other clinical medical data directly related to treatment effect prediction. These information is obtained through medical record collection and clinical evaluation, and is used to comprehensively judge the applicability and expected effect of transplantation treatment, and is an important clinical dimensional basis for precise matching combined with genetic and immune features, and the core value lies in ensuring that the donor selection matches the actual clinical needs of the recipient.

[0031] Before the present scheme is executed, the relevant data of the recipient and the relevant data of each donor in the donor candidate pool can be compared and matched to determine a plurality of target donors adapted to the recipient from a plurality of donors. Specifically, the selection process of the target donor is to compare the matching degree of the multi-dimensional data of the matched recipient and a plurality of candidate donors at the genetic level (such as microbial group similarity, host genetic background), immune characteristics (such as immune response matching degree, inflammation factor compatibility) and clinical indicators (such as disease type, medication history, health status), and dynamically determine the optimal donor combination after evaluation by using the matching algorithm, and the matching result can contain a plurality of adaptive donors to adapt to different treatment stages or individualized adjustment needs.

[0032] Specifically, the fecal flora transplantation donor-recipient dynamic matching method provided in the embodiment can be integrated into a donor-recipient dynamic matching plug-in, and the donor-recipient dynamic matching plug-in can be deployed on any server end that needs to perform fecal flora transplantation donor-recipient matching. After the matching recipient completes the fecal flora transplantation treatment, fecal flora transplantation post-monitoring data at a preset sampling time can be determined at a preset time stage, and the data are stored in a database. When an operation user clicks a response to a donor-recipient matching request in a display interface of the donor-recipient dynamic matching plug-in, the fecal flora transplantation post-monitoring data corresponding to the matching recipient can be obtained from the database. The data reflect the clinical state and microbial colonization effect of the matching recipient at the current stage after receiving the fecal flora transplantation, including the degree of symptom improvement and the survival rate of the transplanted flora, and provide real-time basis for subsequent dynamic adjustment of the treatment scheme.

[0033] S120, inputting the fecal flora transplantation post-monitoring data into a transplantation strategy control model, and outputting a transplantation adjustment strategy for adjusting the matching recipient in a next period.

[0034] The transplantation strategy control model is an intelligent decision system constructed based on an algorithm. The model automatically generates a personalized adjustment scheme for the next treatment stage by analyzing the input postoperative monitoring data of the recipient (such as symptom scores and flora colonization rates) and combining preset treatment rules or optimization strategies obtained through machine learning. The output result of the transplantation strategy control model can involve changes in key treatment parameters such as replacement of donor sources or adjustment of flora transplantation doses, thereby realizing dynamic optimization management of the fecal flora transplantation treatment process.

[0035] The transplantation adjustment strategy is a targeted treatment scheme optimization instruction generated by the transplantation strategy control model. The transplantation adjustment strategy includes at least one of a donor switching strategy and a dose adjustment strategy. The donor switching strategy is a treatment adjustment scheme for replacing a more matched donor selected from a candidate donor library according to a preset algorithm when it is found through evaluation that the flora transplantation effect of the current donor does not meet the expectation during the fecal flora transplantation treatment process. The strategy improves the flora colonization success rate and clinical efficacy by dynamically optimizing the donor-recipient pairing, and is one of the key intervention means of the transplantation adjustment strategy. The dose adjustment strategy is a dynamic regulation method for optimizing the treatment effect by increasing or decreasing the number of flora for subsequent transplantation based on the evaluation result of the recipient fecal flora transplantation post-monitoring data during the fecal flora transplantation treatment process. The purpose of the dose adjustment strategy is to improve the flora colonization effect and the degree of clinical symptom relief by accurately adjusting the input amount of the microbial community, and the dose adjustment strategy is an important part of the individualized adjustment of the transplantation scheme.

[0036] In the present embodiment, the next period refers to the next preset treatment cycle or observation window immediately following the current monitoring time point in the fecal flora transplantation treatment process, which is pre-set by a clinical treatment plan or model parameter, used to implement the transplantation adjustment strategy formulated based on the current evaluation results, and is the key time node for implementing treatment intervention in the dynamic matching process.

[0037] Specifically, based on the obtained post-fecal flora transplantation monitoring data, the post-fecal flora transplantation monitoring data of the to-be-matched recipient can be input into the transplantation strategy control model, and the model generates a targeted transplantation adjustment strategy for guiding the next stage of treatment through algorithm operation. The output result includes two core regulation modes: one is to replace the source of fecal flora according to the donor flora fitness, that is, the donor switching strategy; the other is to adjust the flora delivery amount of subsequent transplantation, that is, the dose adjustment strategy. The system will automatically select one of the two strategies or use both strategies at the same time according to the model evaluation results, so as to realize the individualized dynamic optimization of the treatment plan.

[0038] For example, after integrating the donor-recipient dynamic matching plug-in in the fecal flora transplantation treatment system of a certain hospital, the doctor can automatically record the symptom score and flora colonization rate of patient A to the hospital database according to the preset period (such as the 3rd day, 5th day, 7th day, 14th day, etc. after the operation) after completing the first fecal flora transplantation operation for patient A. On the 4th day after the operation, when the attending physician clicks the "intelligent evaluation" button in the electronic medical record system, the plug-in calls the data of patient A in the 3rd week, that is, the post-fecal flora transplantation monitoring data. Subsequently, these post-fecal flora transplantation monitoring data can be automatically input into the transplantation strategy control model, and the transplantation strategy control model analyzes and finds that the flora functional gene of donor M has a matching degree of only 68% with the patient (lower than the 75% threshold), immediately generates the strategy of "switching to the alternative donor N with a matching degree of 82%", and at the same time detects that the patient's intestinal flora load is insufficient, and adds the suggestion of "increasing the transplantation dose by 20%"; after the medical team reviews, the system automatically schedules the transplantation treatment of 1.2 times the dose of donor N, and sets a new round of postoperative monitoring plan, realizing the whole-process closed-loop management from data collection, intelligent analysis to treatment optimization.

[0039] The technical scheme of the embodiment of the present application responds to a donor-recipient matching request to obtain post-fecal flora transplantation monitoring data of a to-be-matched recipient at a current preset sampling time, wherein the post-fecal flora transplantation monitoring data includes recipient symptom scores and flora colonization rates, the post-fecal flora transplantation monitoring data is determined after a to-be-matched recipient completes a fecal flora transplantation operation based on the fecal flora of any one target donor, and the target donor is determined by adaptively matching the genetic feature information, immune feature information and clinical feature information of the to-be-matched recipient and a plurality of candidate donors; further, the post-fecal flora transplantation monitoring data is input into a transplantation strategy control model to output a transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period; wherein the transplantation adjustment strategy includes at least one of a donor switching strategy and a dose adjustment strategy. The technical scheme provided in this embodiment is based on a dynamic feedback mechanism of real-time monitoring data after transplantation, continuously optimizes donor selection and dose adjustment through a transplantation strategy control model, solves the pain point that a static matching cannot adapt to dynamic changes after transplantation, in addition, integrates multi-modal data such as genetic features, immune features and clinical features, breaks through the limitation of single microbiome data; through multi-modal data integration, comprehensive compatibility evaluation is realized, and with the help of a dynamic feedback mechanism, a treatment scheme is continuously optimized, which significantly improves the precision and adaptability of FMT treatment.

[0040] Figure 2 A fecal flora transplantation donor-recipient dynamic matching method provided by the embodiment of the present application is shown in the figure, which is based on the foregoing embodiment and details the determination method of the target donor. The specific implementation can be referred to the technical scheme of the embodiment. The same or corresponding technical terms as the above embodiments are not repeated here.

[0041] As Figure 2 shown, the method specifically includes the following steps:

[0042] S210, obtaining recipient sample data corresponding to the to-be-matched recipient and donor sample data corresponding to the plurality of candidate donors.

[0043] The recipient sample data refers to a set of multi-dimensional biomedical data collected from the recipient to be matched for evaluating the suitability of fecal microbiota transplantation. The recipient sample data includes recipient fecal microbiota microbial gene data, recipient immune state data, and recipient clinical metadata. The recipient fecal microbiota microbial gene data refers to the microbial genomic information obtained from the fecal sample of the recipient to be matched by high-throughput sequencing and other technologies. It mainly includes genetic characteristic data such as species composition, functional gene distribution, and microbial diversity of the recipient's intestinal flora. These data objectively reflect the structural characteristics and metabolic potential of the recipient's intestinal microecosystem. The recipient immune state data refers to a set of comprehensive indicators reflecting the functional state and response characteristics of the immune system of the recipient to be matched, obtained through laboratory detection and clinical evaluation. It mainly includes immunological parameters such as immune cell composition and activity, cytokine levels, inflammatory marker concentrations, and autoantibody profiles. These data are used to evaluate the potential acceptance or rejection risk of the recipient's immune system to the transplanted bacterial flora. The recipient clinical metadata refers to a set of clinical background and medical history information related to the recipient to be matched, including but not limited to structured clinical data such as the patient's medical history, medication records, allergy history, physiological and biochemical indicators, and imaging examination results. These metadata provide necessary clinical background for evaluating the patient's overall health status, disease progression, and treatment suitability.

[0044] The candidate donor refers to a potential fecal microbiota donor who has passed the preliminary screening and meets the basic donation conditions, but has not yet completed the matching evaluation with a specific recipient. These donors have completed basic screening including microbial gene detection, immune state evaluation, and clinical health examination, and their sample data is included in the donor database as a candidate matching object. The donor sample data refers to a set of multi-dimensional biomedical data collected from the candidate donor for evaluating the suitability of the candidate donor as a fecal microbiota donor. The donor sample data includes donor fecal microbiota microbial gene data, donor immune state data, and recipient clinical metadata. The donor fecal microbiota microbial gene data refers to the microbial genomic information obtained from the fecal sample of a certain candidate donor by high-throughput sequencing and other technologies. The donor immune state data refers to a set of comprehensive indicators reflecting the functional state and response characteristics of the immune system of a certain candidate donor, obtained through laboratory detection and clinical evaluation. The donor clinical metadata refers to a set of clinical background and medical history information related to a certain candidate donor. In this embodiment, each potential donor corresponds to a set of donor sample data.

[0045] In the present embodiment, two types of key data sets can be retrieved from the medical database: one type is the complete biological and clinical data set of the recipient to be matched (including its gut flora gene characteristics, immune system state indicators, and comprehensive clinical background information), i.e., the recipient sample data; the other type is the same dimension data set of the candidate donors after preliminary screening, i.e., the donor sample data corresponding to each candidate donor. This data collection step establishes an analysis foundation containing microbiological, immunological, and clinical medical multi-dimensional information for subsequent precise matching of donors and recipients and is a prerequisite for intelligent matching.

[0046] S220, performing feature extraction processing on the recipient sample data to obtain recipient gene feature information, recipient immune feature information, and recipient clinical feature information.

[0047] In the present embodiment, the recipient gene feature information refers to a set of key genetic indicators representing the composition and functional characteristics of the recipient's gut flora obtained by performing feature extraction processing on the recipient's fecal microbial gene data. The recipient immune feature information refers to a set of key immunological indicators representing the functional characteristics and reaction patterns of the recipient's immune system obtained by performing feature extraction processing on the recipient's immune state data. The recipient clinical feature information refers to a set of key clinical indicators representing the overall health status and disease characteristics of the recipient obtained by performing feature extraction processing on the recipient's clinical metadata.

[0048] Specifically, the recipient fecal microbial gene data includes metagenomic sequencing data, species composition types of the recipient's fecal sample, and relative abundance data of each species, and the recipient gene feature information includes microbial species relative abundance features, microbial co-occurrence network graphs, and functional gene prediction feature information.

[0049] The metagenomic sequencing data refers to genetic information obtained by high-throughput sequencing of the bacterial 16S ribosomal RNA gene in the recipient's fecal sample. This gene serves as a marker gene for bacterial classification, and its sequence variation can accurately reflect the composition and abundance distribution of different bacterial species in the microbial community. The species composition type refers to the taxonomic information and community structure characteristics of the microbial species present in the recipient's or donor's fecal sample obtained by bioinformatics analysis of the metagenomic sequencing data, including the distribution of different classification levels such as bacterial phylum, class, order, family, genus, and species. The relative abundance data of each species refers to quantitative indicators of the proportion of each microbial species in the overall flora in the recipient's or donor's fecal sample obtained by bioinformatics analysis of the metagenomic sequencing data. This data accurately represents the quantitative distribution of different classification levels (such as genus and species) of microorganisms in the community in the form of percentages or proportions.

[0050] On this basis, the specific implementation steps of the donor gene feature information obtained by feature extraction processing of the donor fecal microbial gene data can include:

[0051] S01, the species composition type of the recipient fecal sample and the relative abundance data of each species are processed by feature vectorization to obtain the microbial species relative abundance feature.

[0052] Among them, the microbial species relative abundance feature refers to the quantitative representation obtained by structuring or dimensionality reduction processing of the relative abundance data of different microbial species in the fecal sample by mathematical or statistical methods, which reflects the composition distribution pattern of the microbial community, usually in the form of numerical vector or matrix.

[0053] In this embodiment, all species composition types and their relative abundance information detected in the recipient fecal sample can be converted into structured numerical representation, and the original species abundance data can be integrated into standardized feature vectors that can be used for calculation or modeling by mathematical conversion or statistical methods, that is, the microbial species relative abundance feature can be obtained. In this way, by converting the complex microbial community composition information into low-dimensional numerical features that can reflect the overall distribution pattern, subsequent machine learning analysis or pattern recognition is facilitated.

[0054] S02, the species composition type of the recipient fecal sample and the relative abundance data of each species are processed by network graph construction by co-occurrence network analysis tool to obtain the microbial co-occurrence network graph.

[0055] Among them, the co-occurrence network analysis tool refers to a class of calculation software or algorithm platform specially used for identifying and visualizing the coexistence relationship between microbial species, which detects the co-occurrence pattern of different species in multiple samples by statistical methods (such as correlation analysis, mutual information calculation, etc.), and constructs the species interaction network according to the preset significance threshold, and finally generates a visual network graph containing topological structure parameters.

[0056] Among them, the microbial co-occurrence network graph refers to a visual network structure used to show the coexistence relationship pattern between different species in the microbial community. The nodes in the microbial co-occurrence network graph represent different species, and the connection edges between the nodes represent the significant correlation between the species. It can be understood that the edge connecting the nodes represents a statistically significant positive correlation (symbiosis) or negative correlation (competition) between the species, and the weight of the edge usually reflects the strength of the correlation. This network graph can intuitively present the ecological interaction characteristics and community structure stability of the microbial community.

[0057] In the present embodiment, a special co-occurrence network analysis tool can be used to screen out species pairs with significant co-occurrence relationship based on the species composition types of the microbial species in multiple recipient fecal samples and the relative abundance data of each species by calculating the statistical correlation (such as Spearman rank correlation, SparCC, etc.) between species, and convert these relationships into a network topology structure composed of nodes (representing species) and edges (representing significant correlation), thereby forming a microbial co-occurrence network diagram that can intuitively display the interaction relationship between species in the microbial community.

[0058] S03, performing gene function prediction analysis on the metagenomic sequencing data by using a PICRUSt2 function prediction tool to obtain function gene prediction results, and performing feature vectorization processing on the function gene prediction results to obtain function gene prediction feature information.

[0059] PICRUSt2 function prediction tool is a bioinformatics software for predicting the function of microbial communities based on metagenomic sequencing data. The function gene prediction result refers to the functional feature data of the microbial community obtained by indirect inference on the metagenomic sequencing data, which is essentially based on the correspondence between species classification and functional genes in the known microbial genome database. The function gene prediction feature information refers to the quantitative representation data obtained by mathematical conversion and structured processing of the function gene prediction results, which converts the original prediction results into computable numerical features through dimensionality reduction or feature extraction methods.

[0060] In the present embodiment, the PICRUSt2 tool can be used to infer the potential functional genomic composition of the microbial community based on 16S rRNA gene sequence data to obtain function gene prediction results. Further, these function gene prediction results are converted into structured and quantifiable numerical features by mathematical methods, thereby obtaining function gene prediction feature information. In this way, the complex gene function prediction data is converted into a standardized feature representation suitable for machine learning algorithm processing, thereby realizing the quantitative analysis and comparison of the functional features of the microbial community.

[0061] Optionally, the specific implementation of the method for obtaining the recipient immune feature information by performing feature extraction processing on the recipient immune state data comprises: performing feature vectorization processing on the recipient immune state data to obtain the recipient immune feature information.

[0062] In this embodiment, the original recipient immune status data (such as cytokine levels, immune cell counts, or antibody titers, etc.) can be converted into structured numerical vectors through mathematical conversion or standardization methods to obtain recipient immune feature information. This processing can integrate multi-dimensional immune indicators into machine-readable quantitative features, thereby converting complex immune status information into a digital representation form that can be used for computational analysis and modeling, facilitating subsequent immune feature analysis and pattern recognition.

[0063] Optionally, the specific implementation of the feature extraction processing of the recipient clinical metadata to obtain the recipient clinical feature information can include: performing feature vectorization processing on the recipient clinical metadata based on a bidirectional encoder representation transformation model to obtain a clinical data feature vector.

[0064] The bidirectional encoder representation transformation model is a natural language processing architecture based on deep learning, and its core feature is to capture the context bidirectional semantic information of each word in the text data through a bidirectional Transformer structure. Through pre-training and fine-tuning two stages, the model learns to convert input text sequences into dense vector representations containing rich semantic features, thereby realizing deep semantic coding and feature extraction of text data, and is particularly suitable for processing text feature representation tasks that require understanding of global context relationships.

[0065] The recipient clinical metadata includes at least one of recipient attribute information, clinical symptom description information, recipient disease type, past medical history information, family medical history information, recent medication information, lifestyle information, human leukocyte antigen typing, and blood type information. The recipient attribute information refers to the basic demographic characteristics of the patient; the clinical symptom description information is a text record of the patient's subjective discomfort and objective signs; the recipient disease type represents the disease classification diagnosed by the patient; the past medical history information contains the past diagnosed diseases and treatment of the patient; the family medical history information records the genetic disease history of the direct relatives; the recent medication information reflects the current or recent medication information of the patient; the lifestyle information covers daily behavior habits such as diet and exercise; the human leukocyte antigen typing is the detection result of the immune-related genotype; and the blood type information refers to the ABO and Rh red cell antigen typing results.

[0066] In this embodiment, the bidirectional encoder representation transformation model can be used to convert clinical metadata containing text and numerical types (such as symptom description, medical history record, etc.) into low-dimensional dense numerical vector representations through a deep learning model to obtain a clinical data feature vector. The model has semantic understanding ability obtained through pre-training, which can capture the context association and potential medical features in the clinical text, thereby generating a feature vector containing rich clinical semantic information, providing structured numerical representation for subsequent medical data analysis.

[0067] S230, feature extraction processing is performed on each group of donor sample data to obtain donor gene feature information, donor immune feature information and donor clinical feature information corresponding to each candidate donor.

[0068] In this embodiment, for each group of donor sample data, the same feature extraction method as the recipient sample data can be used for feature extraction processing, so that donor gene feature information, donor immune feature information and donor clinical feature information corresponding to each group of donor sample data can be obtained.

[0069] On the basis of obtaining the recipient gene feature information, the recipient immune feature information and the recipient clinical feature information, and the donor gene feature information, the donor immune feature information and the donor clinical feature information corresponding to each donor, these data contents can be stored in a pre-set storage space.

[0070] S240, obtaining recipient sample feature data corresponding to the to-be-matched recipient and donor sample feature data corresponding to a plurality of donors.

[0071] The recipient sample feature data includes disease type and severity identifier, recipient gene feature information, recipient immune feature information and recipient clinical feature information; the donor sample feature data includes donor gene feature information, donor immune feature information and donor clinical feature information. The disease type and severity identifier refers to the standardized classification code of the disease suffered by the patient determined by medical diagnosis and the quantitative evaluation index of the current development stage and clinical symptom severity of the disease.

[0072] In this embodiment, when it is necessary to determine a plurality of target donors suitable for the recipient, the recipient sample feature data corresponding to the to-be-matched recipient and the donor sample feature data corresponding to a plurality of donors can be called from the pre-set storage space.

[0073] S250, for each candidate donor, the recipient sample feature data and the donor sample feature data corresponding to the current candidate donor are matched by the pre-trained donor-recipient matching model to obtain microbial compatibility attribute and clinical adaptation attribute.

[0074] The donor-recipient matching model refers to an intelligent matching system constructed by a machine learning algorithm. Based on the multidimensional biomedical feature data of the recipient and the donor (including gene features, immune features and clinical features, etc.), the matching degree of both parties in microbial compatibility and clinical adaptation is calculated, so as to quantitatively evaluate the transplantation suitability between the recipient and each potential donor, and provide a predictive analysis model for optimal donor screening decision support for fecal flora transplantation.

[0075] The microbial compatibility attribute refers to a quantitative index representing the matching degree of the microbial community between the donor and the recipient, which comprehensively evaluates the similarity and complementarity of the fecal microbial species composition, functional genes and ecological network of the two parties, and reflects the potential adaptability of the donor flora in the recipient in terms of intestinal colonization and functional integration.

[0076] The clinical adaptation attribute refers to a quantitative index representing the matching degree of the recipient and the donor in the clinical feature dimension, which comprehensively evaluates the compatibility and complementarity of the two parties in terms of disease type, physiological parameters, medication history, living habits and other clinical medical characteristics, and reflects the potential treatment adaptation degree of the donor flora in improving the specific clinical symptoms of the recipient.

[0077] For each candidate donor, the matching process of the recipient with each candidate donor is consistent. In order to clearly introduce the technical solution, one of the candidate donors can be taken as an example for introduction. The current candidate donor refers to the candidate donor being matched and processed. In the specific application process, the trained and optimized donor-recipient matching model can be used to comprehensively analyze the multi-dimensional data of the microbial characteristics, immune characteristics and clinical characteristics of the recipient to be matched and the current candidate donor by a specific algorithm, so as to output two key matching indexes: the microbial compatibility attribute reflecting the compatibility of the microbial community structure and function of the two parties, and the clinical adaptation attribute evaluating the matching degree of the clinical characteristics and treatment needs of the two parties.

[0078] More specifically, the donor-recipient matching model can include a microbial compatibility evaluation model and a clinical adaptation evaluation model. The recipient gene feature information includes microbial co-occurrence network graph and functional gene prediction feature information. Optionally, based on the pre-trained donor-recipient matching model, the matching calculation of the recipient sample feature data and the corresponding donor sample feature data of the current candidate donor can include the following specific implementation steps:

[0079] S2501, based on the recipient microbial species relative abundance feature in the recipient sample feature data, the recipient microbial co-occurrence network feature information, the donor microbial species relative abundance feature in the donor sample feature data, the donor microbial co-occurrence network feature information and the microbial compatibility evaluation model, the microbial compatibility attribute between the recipient to be processed and the current candidate donor is determined.

[0080] The microbial compatibility evaluation model refers to a calculation model for quantitatively evaluating the compatibility of the microbial community between the recipient and the donor. The model analyzes the microbial species composition, abundance distribution and ecological network characteristics of the two parties, and uses a specific compatibility evaluation algorithm to calculate the matching degree index reflecting the colonization potential and functional integration ability of the donor flora in the recipient intestinal environment, thereby providing a scientific basis for donor screening for fecal flora transplantation.

[0081] In this embodiment, the pre-established microbial compatibility evaluation model can be used to calculate the similarity and complementarity of the recipient and the donor in the microbial species composition (relative abundance characteristics) and ecological interaction relationship (co-occurrence network characteristics) in two dimensions, comprehensively evaluate the matching degree of the two microbial communities in structural characteristics and functional relationship, and finally output a numerical index quantifying the compatibility of the recipient and the donor microbial community, i.e., the microbial compatibility attribute between the recipient to be processed and the current candidate donor.

[0082] Optionally, the specific implementation of determining the microbial compatibility attribute between the recipient to be processed and the current candidate donor by the microbial compatibility evaluation model can include:

[0083] S11, the relative abundance data corresponding to each microbial species in the recipient sample data and the evolutionary relationship coding information corresponding to each species are configured as node parameter information of the corresponding species in the recipient microbial co-occurrence network graph, and a recipient characteristic microbial co-occurrence network graph is obtained.

[0084] The evolutionary relationship coding information refers to the conversion of the evolutionary distance and genetic relationship between microbial species into a numerical representation data form. It can be based on multiple alignment and phylogenetic tree construction of 16S rRNA gene sequences or other conserved gene sequences, and a specific coding method is used to quantitatively represent the relative position and genetic correlation degree of different species in the evolutionary lineage.

[0085] The recipient characteristic microbial co-occurrence network graph refers to an enhanced network structure that integrates the relative abundance data and evolutionary relationship coding information of the microbial species in the recipient sample. The node represents a specific microbial species and carries the abundance characteristics and phylogenetic position information of the species, and the edge represents the statistically verified ecological interaction relationship between species. This network graph provides a comprehensive topological representation containing species quantitative composition and evolutionary background for subsequent graph neural network processing by integrating multi-dimensional microbial feature data.

[0086] In this embodiment, the relative abundance data corresponding to each microbial species detected in the recipient sample and the evolutionary relationship coding information corresponding to each species are collectively encoded as feature vectors of the recipient microbial co-occurrence network graph. These enhanced feature vectors are assigned to the corresponding species nodes in the co-occurrence network, thereby constructing a network representation that contains both species ecological association and integrates quantitative composition and evolutionary characteristics, i.e., the recipient characteristic microbial co-occurrence network graph. This processing realizes the structured integration and network representation of multi-dimensional microbial community data.

[0087] S12, the relative abundance data corresponding to each microbial species in the donor sample data and the evolutionary relationship coding information corresponding to each species are configured as node parameter information of the corresponding species of the donor microbial co-occurrence network graph to obtain a donor characteristic microbial co-occurrence network graph.

[0088] Specifically, the determination method of the donor characteristic microbial co-occurrence network graph is consistent with the determination method of the recipient characteristic microbial co-occurrence network graph, and the specific implementation manner can be referred to S11, which will not be repeated here.

[0089] S13, the recipient characteristic microbial co-occurrence network graph and the donor characteristic microbial co-occurrence network graph are respectively input into a pre-trained graph neural network to obtain recipient microbial feature embedding and donor microbial feature embedding.

[0090] The microbial feature embedding refers to a low-dimensional vector representation obtained after deep feature learning of the characteristic microbial co-occurrence network graph. The vector captures the comprehensive characteristics of the microbial community in terms of species composition, abundance distribution, evolutionary relationship and ecological interaction through nonlinear transformation of the neural network, and compresses the complex network topology and node attribute information into a dense numerical vector with rich semantics, providing standardized feature representation for subsequent compatibility comparison. The recipient microbial feature embedding refers to the microbial feature embedding obtained by feature extraction of the recipient characteristic microbial co-occurrence network graph; the donor microbial feature embedding refers to the microbial feature embedding obtained by feature extraction of the donor characteristic microbial co-occurrence network graph.

[0091] In this embodiment, the trained graph neural network can be used to perform deep feature extraction and dimension reduction processing on the recipient characteristic microbial co-occurrence network graph to obtain the recipient microbial feature embedding. Similarly, the graph neural network is used to perform deep feature extraction and dimension reduction processing on the donor characteristic microbial co-occurrence network graph to obtain the donor microbial feature embedding. In this way, the complex network structure data is converted into a low-dimensional dense vector representation that can fully represent the characteristics of the respective microbial community through the hierarchical information transmission and aggregation mechanism of the network.

[0092] S14, based on the similarity attribute between the recipient microbial feature embedding and the donor microbial feature embedding, determining the microbial compatibility attribute between the to-be-matched recipient and the current candidate donor.

[0093] In this embodiment, the distance or similarity index (such as cosine similarity or Euclidean distance) in the embedding space between the recipient microbial feature embedding and the donor microbial feature embedding can be calculated to quantitatively evaluate the matching degree of the microbial communities of both parties in terms of composition structure, functional potential and ecological relationship, so as to obtain a microbial compatibility attribute that can objectively reflect the colonization and functional integration potential of the current candidate donor bacterial community in the intestinal environment of the to-be-matched recipient.

[0094] S2502, determine the clinical adaptation attribute between the to-be-processed recipient and the current candidate donor based on the recipient functional gene prediction characteristic information, the recipient immune characteristic information, the recipient clinical characteristic information, the donor functional gene prediction characteristic information, the donor immune characteristic information, the donor clinical characteristic information in the donor sample characteristic data, and a clinical adaptation evaluation model.

[0095] The clinical adaptation evaluation model refers to a calculation model for evaluating the matching degree of the recipient and the donor in the clinical characteristic dimension. The model analyzes the multi-dimensional clinical data of both parties in disease characteristics, immune state, medication history, and living habits, and calculates the matching degree index reflecting the improvement potential of the donor flora on the specific clinical symptoms of the recipient by using a specific matching algorithm, thereby providing a clinical adaptation evaluation basis for the fecal flora transplantation.

[0096] In this embodiment, the matching degree of the to-be-matched recipient and the current candidate donor in the functional gene composition, immune state, and clinical characteristics can be comprehensively analyzed by the clinical adaptation evaluation model, and the potential improvement effect and treatment adaptation of the donor flora on the specific disease symptoms of the recipient can be quantitatively evaluated, and finally a quantitative index reflecting the clinical treatment suitability of both parties is output, i.e., the clinical adaptation attribute between the to-be-processed recipient and the current candidate donor.

[0097] Optionally, the specific implementation manner of determining the clinical adaptation attribute between the to-be-processed recipient and the current candidate donor by the clinical adaptation evaluation model can include:

[0098] S21, construct a to-be-processed vector of the recipient based on the recipient functional gene prediction characteristic information, the recipient immune characteristic information, and the recipient clinical characteristic information in the recipient sample data.

[0099] The to-be-processed vector of the recipient refers to a structured numerical vector formed by standardizing and fusing the multi-dimensional medical indicators such as the functional gene prediction characteristics, immune characteristics, and clinical characteristics in the recipient sample data.

[0100] In this embodiment, the feature data of the to-be-matched recipient in the three dimensions of the recipient functional gene prediction characteristic information, the recipient immune characteristic information, and the recipient clinical characteristic information can be standardized and fused to form a unified numerical vector, i.e., the to-be-processed vector of the recipient. The vector converts the multi-source heterogeneous clinical and biological characteristics into machine-processable standardized numerical characteristics by a specific mathematical conversion and dimension reduction method, thereby providing structured input data for subsequent transplantation adaptation prediction.

[0101] S22, construct a donor to-be-processed vector based on the donor functional gene prediction feature information, the donor immune feature information, and the donor clinical feature information in the donor sample data.

[0102] The donor to-be-processed vector refers to a structured feature vector formed by processing and numerical standardization conversion of multi-dimensional biomedical indicators such as functional gene prediction features, immune features, and clinical features in the donor sample data.

[0103] Specifically, the determination manner of the donor to-be-processed vector is consistent with that of the recipient to-be-processed vector, and the specific implementation manner can be referred to S21, which will not be described here.

[0104] S23, input the recipient to-be-processed vector and the donor to-be-processed vector into the pre-trained Transformer model to obtain a postoperative infection risk probability and a rejection reaction risk probability.

[0105] The postoperative infection risk probability refers to a numerical indicator for quantifying the possibility of pathogenic microorganism infection of the recipient after receiving the fecal bacteria transplantation surgery. The rejection reaction risk probability refers to a numerical indicator for quantitatively evaluating the possibility of abnormal immune response of the recipient's immune system to the transplanted fecal bacteria.

[0106] In this embodiment, the trained Transformer deep learning model can be used to jointly analyze and extract deep features of the recipient to-be-processed vector and the donor to-be-processed vector. Through the self-attention mechanism of the model, the complex interaction relationship between the two parties in the functional gene, immune feature, and clinical feature can be captured, and finally the postoperative infection risk probability quantifying the possibility of postoperative infection and the rejection reaction risk probability quantifying the possibility of immune rejection can be output.

[0107] S24, determine the clinical adaptation attribute between the to-be-processed recipient and the current candidate donor based on the postoperative infection risk probability and the rejection reaction risk probability.

[0108] In this embodiment, by comprehensively evaluating the postoperative infection risk probability and the rejection reaction risk probability predicted by the model, two key clinical indicators, a specific decision algorithm or scoring rule is used to convert the two risk probability values into a comprehensive clinical adaptation score, which is the clinical adaptation attribute between the to-be-processed recipient and the current candidate donor. The score quantitatively reflects the overall balance degree of the clinical treatment effect and safety that the current donor flora transplantation to the recipient can achieve.

[0109] S260, determine the matching attribute between the to-be-matched recipient and the current candidate donor based on the disease type information and the severity label information of the to-be-matched recipient, the microbial compatibility attribute, and the clinical adaptation attribute.

[0110] The matching attribute refers to a quantitative index obtained by comprehensively evaluating the microbial compatibility attribute, the clinical adaptation attribute, and the disease severity weight, which reflects the overall matching degree of the to-be-matched recipient and the specific donor in the three dimensions of microbial community compatibility, clinical treatment adaptation, and disease-specific demand, and is the final comprehensive evaluation result for determining the priority of the donor selection.

[0111] In this embodiment, the evaluation weights of the microbial compatibility and the clinical adaptation can be dynamically adjusted according to the disease type information and the severity label information, the evaluation indexes of the two dimensions are comprehensively analyzed through a specific weighted fusion algorithm, and finally a quantitative comprehensive matching score is generated, which is the matching attribute between the to-be-matched recipient and the current candidate donor.

[0112] Optionally, the specific implementation steps of S260 can include:

[0113] S2601, determining the microbial compatibility weight and the clinical adaptation weight based on the disease type information and the severity label information of the to-be-matched recipient.

[0114] The microbial compatibility weight refers to a relative importance coefficient allocated to the microbial compatibility attribute. The clinical adaptation weight refers to a relative importance coefficient allocated to the clinical adaptation attribute.

[0115] In this embodiment, the relative importance proportions of the two key dimensions of the microbial compatibility evaluation and the clinical adaptation evaluation in the final matching decision can be dynamically calculated and determined according to the disease type information and the severity label information of the to-be-matched recipient through a preset weight allocation rule, that is, the microbial compatibility weight and the clinical adaptation weight, so as to realize the differentiated regulation of the contribution degrees of various evaluation factors under different disease states.

[0116] Optionally, the specific implementation manner of determining the microbial compatibility weight and the clinical adaptation weight based on the disease type information and the severity label information of the to-be-matched recipient can include:

[0117] If it is determined that the to-be-matched recipient is a primary patient based on the disease type information and the severity label information of the to-be-matched recipient, the microbial compatibility weight is determined from a first preset numerical range, and the clinical adaptation weight is determined from a second preset numerical range; if it is determined that the to-be-matched recipient is a refractory patient based on the disease type information and the severity label information of the to-be-matched recipient, the microbial compatibility weight is determined from a third preset numerical range, and the clinical adaptation weight is determined from a fourth preset numerical range.

[0118] The numerical values of the first preset numerical range are all greater than the numerical values of the second preset numerical range, and the numerical values of the third preset numerical range are all less than the numerical values of the fourth preset numerical range.

[0119] In this embodiment, a differentiated weight allocation strategy can be adopted, that is, the importance weights of microbial compatibility and clinical adaptability are dynamically adjusted according to the disease type information and severity label information (initial or refractory): for initial patients, a higher weight (a value in the first preset numerical range > a value in the second preset numerical range) can be given to microbial compatibility, emphasizing the basic role of flora matching; and for refractory patients, more attention is paid to clinical adaptability (a value in the fourth preset numerical range > a value in the third preset numerical range), highlighting the accuracy of treatment effect. This differentiated design embodies the personalized matching strategy for patients at different stages of the disease.

[0120] S2602, based on the microbial compatibility weight and the clinical adaptability weight, performing weighted operation on the microbial compatibility attribute and the clinical adaptability attribute between the to-be-matched recipient and the current candidate donor.

[0121] In this embodiment, the microbial compatibility score and the clinical adaptability score are multiplied by the respective weight coefficients and then added, and through this weighted summation calculation method, the evaluation results of the two dimensions are integrated according to their importance proportions, and finally a comprehensive matching attribute score is generated, which quantitatively reflects the overall adaptability of the donor and the recipient in terms of microbial community and clinical characteristics.

[0122] S270, based on the matching attributes between the to-be-matched recipient and each donor, determining a plurality of target donors adapted to the to-be-matched recipient.

[0123] In this embodiment, according to the pre-set matching threshold or sorting rule, the matching attributes between all candidate donors and the recipient are screened or sorted, and several candidate donors whose matching degrees meet the required standard or ranking are selected as the final recommended transplantation objects, which are the plurality of target donors adapted to the to-be-matched recipient. These target donors show the best matching degree with the recipient in terms of microbial compatibility and clinical adaptability.

[0124] S280, in response to the recipient-donor matching request, obtaining fecal flora transplantation post-monitoring data of the to-be-matched recipient at a current pre-set sampling time.

[0125] The fecal flora transplantation post-monitoring data includes recipient symptom score and flora colonization rate, and the fecal flora transplantation post-monitoring data is determined after the to-be-matched recipient completes the fecal flora transplantation surgery based on the fecal flora of any target donor.

[0126] S290, inputting the fecal flora transplantation post-monitoring data into the transplantation strategy control model, and outputting a transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period.

[0127] The transplantation adjustment strategy includes at least one of a donor switching strategy and a dose adjustment strategy.

[0128] The technical solution of the embodiment of the application, in determining a plurality of target donors, acquires the sample feature data of the recipient corresponding to the to-be-matched recipient and the sample feature data of a plurality of candidate donors corresponding to the candidate donors; wherein the sample feature data of the recipient includes disease type and severity identification, recipient genetic feature information, recipient immune feature information, and recipient clinical feature information; the sample feature data of the donor includes donor genetic feature information, donor immune feature information, and donor clinical feature information; for each candidate donor, the sample feature data of the recipient and the sample feature data of the current candidate donor corresponding to the candidate donor are matched and calculated based on a pre-trained donor-recipient matching model to obtain microbial compatibility attributes and clinical adaptation attributes; based on the disease type information and severity label information of the to-be-matched recipient, the microbial compatibility attributes, and the clinical adaptation attributes, the matching attributes between the to-be-matched recipient and the current candidate donor are determined; based on the matching attributes between the to-be-matched recipient and each candidate donor, a plurality of target donors adapted to the to-be-matched recipient are determined. The technical solution provided in the embodiment integrates the quantitative evaluation of microbial compatibility and clinical adaptability in two dimensions, and combines disease-specific weight adjustment to achieve scientific and objective donor screening; a multi-parameter comprehensive scoring mechanism is used to ensure the comprehensiveness and accuracy of the matching result; at the same time, the selection scheme of multiple adaptive donors is retained, which not only provides flexible and diverse transplantation options for clinical treatment, but also can dynamically adjust the donor strategy according to the postoperative reaction of the patient, significantly improving the individualization and success rate of fecal flora transplantation treatment.

[0129] Figure 3 A schematic diagram of the fecal flora transplantation donor-recipient dynamic matching method provided in the embodiment of the application is further refined on the basis of the foregoing embodiment, and the specific implementation process of S120 can be referred to the technical solution of the embodiment. Wherein, the same or corresponding technical terms as the above embodiments are not repeated here.

[0130] As Figure 3 shown, the method specifically includes the following steps:

[0131] S310, in response to a donor-recipient matching request, acquiring fecal flora transplantation post-monitoring data of a to-be-matched recipient at a current preset sampling time.

[0132] Wherein, the fecal flora transplantation post-monitoring data includes recipient symptom score and flora colonization rate, and the fecal flora transplantation post-monitoring data is determined after the to-be-matched recipient completes the fecal flora transplantation surgery based on the fecal flora of any one target donor; the target donor is a plurality of donors determined based on the genetic feature information, immune feature information, and clinical feature information of the to-be-matched recipient and a plurality of candidate donors.

[0133] In the embodiment, the transplantation strategy control model includes an agent decision network and a virtual environment network, and the agent decision network includes an action value mapping layer and an action decision layer. Specifically, the transplantation strategy control model can adopt a decision system constructed by a reinforcement learning framework. The agent decision network refers to evaluating the value parameters of each potential transplantation strategy through the action value mapping layer, and selecting the optimal execution action by the action decision layer; the virtual environment network refers to a digital simulation system simulating the physiological response after fecal flora transplantation. The virtual environment network receives the action instruction of the decision network, simulates the clinical effect, and feeds back the state update and reward signal, so as to realize the iterative optimization of the transplantation strategy.

[0134] S320, monitoring data after fecal flora transplantation is taken as initial state information, and the initial state information is taken as current state information.

[0135] The current state information refers to dynamic data representing the comprehensive clinical state and microbial colonization of the recipient at a certain decision time in the process of strategy optimization iteration.

[0136] In the embodiment, the monitoring data after fecal flora transplantation is standardized to form a structured data representation. This information can be taken as the starting input of the reinforcement learning process, and completely represents the immediate clinical state and microbial colonization of the recipient after receiving the current donor fecal flora transplantation, thereby providing a basic state benchmark for subsequent strategy optimization.

[0137] S330, inputting the current state information into the action value mapping layer to obtain a value parameter corresponding to at least one preset execution action.

[0138] The preset execution action refers to a set of optional operations defined in advance in the process of transplantation strategy optimization. More specifically, the at least one preset execution action includes switching to each donor in a plurality of target donors, and a random combination with at least one dose adjustment method. In the decision process, the set of optional treatment schemes defined in advance includes the basic operation of switching to a certain target donor alone, and also includes the composite treatment scheme formed by combining the donor switching with different dose adjustment strategies (such as increment, decrement or maintenance), so as to ensure that the decision space covers all possible clinical intervention methods. The value parameter refers to a numerical index quantitatively representing the expected effect of each preset execution action under the current state, which comprehensively evaluates the potential benefits of the candidate treatment scheme in improving clinical symptoms, maintaining flora balance and reducing adverse reactions, etc.

[0139] In the embodiment, the initial state information characterizing the current clinical state of the recipient is input as the current state information to the agent decision network including an action value mapping layer, through which the built-in algorithm calculates the value parameters of each preset execution action in the current state, which quantitatively reflect the balance between the therapeutic effect and the risk of each candidate action.

[0140] S340, input each value parameter to the action decision layer to obtain a target value parameter satisfying a preset condition, and determine the preset execution action corresponding to the target value parameter as the current execution action.

[0141] Among them, the current execution action refers to the optimal treatment plan selected from the preset execution action set through the agent decision network in the decision process. It represents the donor switching or dose adjustment strategy that can best optimize the balance between therapeutic effect and safety in the current state, and is the immediate clinical intervention measure recommended by the system for implementation.

[0142] In the embodiment, each value parameter can be input to the action decision layer, and the value parameters of all preset execution actions can be filtered and compared by the action decision layer to select the optimal score that meets the optimization standard (such as the highest value or the lowest risk). The score is the target value parameter, so that the preset execution action corresponding to the target value parameter can be determined as the current optimal decision as the current execution action.

[0143] S350, after simulating the execution of the current execution action in the virtual environment network, output the action reward value and the next state information corresponding to the current execution action.

[0144] Among them, the action reward value refers to a numerical evaluation index quantitatively representing the expected effect of the current execution action (donor switching or dose adjustment strategy). The action reward value is determined according to the clinical remission prediction rate, adverse reaction prediction rate, and bacterial diversity index prediction value. The action reward value comprehensively considers the comprehensive benefits of the current execution action in improving clinical symptoms (clinical remission prediction rate), reducing adverse reactions (adverse reaction prediction rate), and optimizing microbial balance (bacterial diversity index prediction value) in three dimensions, as a key feedback signal for strategy optimization in the reinforcement learning process.

[0145] Among them, the next state information refers to the clinical state data representation of the recipient that may be presented in the next period after simulating the execution of the current execution action in the virtual environment network. The information includes updated symptom scores and bacterial colonization rates and other key indicators calculated by simulation, reflecting the expected changes in the recipient's state after taking the current transplantation adjustment strategy.

[0146] In the present embodiment, after the virtual deduction of the current execution action (i.e. donor switching or dose adjustment) through the virtual environment network, two key outputs are generated: the action reward value quantitatively evaluating the overall benefits of the strategy in terms of symptom improvement, safety and flora balance; and the next state information predicting the new clinical state that the recipient will reach after implementing the strategy, both of which together constitute the feedback signal for the iterative optimization of the strategy.

[0147] S360, update the next state information as the current state information, and repeat the steps of determining the value parameter corresponding to the at least one preset execution action, determining the current execution action, and determining the action reward value and the next state information corresponding to the current execution action, until the action reward value obtained by repeating the steps for a plurality of times meets a preset condition, and the current execution action is taken as the transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period.

[0148] Specifically, through the continuous state update and action evaluation loop, i.e. taking the predicted next state as the new current state, recalculating the action value and selecting the optimal action, the effects of different treatment strategies are simulated multiple times in the virtual environment, and when the cumulative action reward value reaches the preset optimal standard (such as convergence or maximization), the iteration is terminated, and finally the transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period is output. This process realizes the dynamic optimization and verification of the treatment scheme in the simulation environment.

[0149] The technical scheme of the embodiment of the present application determines the fecal flora transplantation post-monitoring data as initial state information when determining the transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period, and determines the initial state information as current state information, and then inputs the current state information into the action value mapping layer to obtain a value parameter corresponding to at least one preset execution action, wherein the at least one preset execution action includes: switching to each donor in a plurality of target donors and a random combination with at least one dose adjustment mode; further, input each value parameter into the action decision layer to obtain a target value parameter that meets a preset condition, determine the preset execution action corresponding to the target value parameter as a current execution action, and then output an action reward value corresponding to the current execution action and next state information after simulating the execution of the current execution action in a virtual environment network, wherein the action reward value is determined according to a clinical remission prediction rate, an adverse reaction prediction rate and a flora diversity index prediction value; and then update the next state information as the current state information, and repeat the steps of determining the value parameter corresponding to the at least one preset execution action, determining the current execution action and determining the action reward value corresponding to the current execution action and the next state information until the action reward value obtained by repeating a plurality of times meets the preset condition, and the current execution action is used as the transplantation adjustment strategy for adjusting the to-be-matched recipient in the next period. The technical scheme provided by the embodiment first verifies the clinical effect of different treatment schemes through multiple rounds of virtual deduction, significantly reducing the actual treatment risk; secondly, dynamically integrates multi-dimensional indexes such as symptom improvement, safety and flora balance to ensure the scientificity and comprehensiveness of strategy selection; finally, the iterative optimization mechanism can adaptively adjust the donor selection and dose scheme, providing personalized treatment paths for complex cases, and overall improving the accuracy and efficacy stability of fecal flora transplantation. This method deeply integrates the prediction ability of artificial intelligence with the clinical decision-making demand, creating a new paradigm for intelligent fecal flora transplantation treatment.

[0150] Figure 4 The structure diagram of the fecal flora transplantation donor-recipient dynamic matching device provided by the embodiment of the present application, the device comprises: a monitoring data acquisition module 410 and an adjustment strategy determination module 420.

[0151] The monitoring data acquisition module 410 is configured to acquire the fecal flora transplantation post-monitoring data of the to-be-matched recipient at a current preset sampling time in response to a donor-recipient matching request; wherein the fecal flora transplantation post-monitoring data includes a recipient symptom score and a flora colonization rate, the fecal flora transplantation post-monitoring data is determined after the to-be-matched recipient completes a fecal flora transplantation operation based on the fecal flora of any one target donor, and the target donor is determined by adaptively matching the genetic feature information, immune feature information and clinical feature information of the to-be-matched recipient and a plurality of candidate donors.

[0152] The adjustment strategy determination module 420 is configured to input the post-fecal bacteria transplantation monitoring data into a transplantation strategy control model, and output a transplantation adjustment strategy for adjusting the transplantation of the to-be-matched recipient in a next period; wherein the transplantation adjustment strategy comprises at least one of a donor switching strategy and a dose adjustment strategy.

[0153] The technical solution of the embodiment of the present application responds to a donor-recipient matching request, acquires post-fecal bacteria transplantation monitoring data of a to-be-matched recipient at a current preset sampling time, wherein the post-fecal bacteria transplantation monitoring data comprises a recipient symptom score and a flora colonization rate, and the post-fecal bacteria transplantation monitoring data is determined after a to-be-matched recipient completes a fecal bacteria transplantation operation based on a fecal flora of any one target donor; the target donor is determined by adaptively matching gene feature information, immune feature information and clinical feature information of the to-be-matched recipient and a plurality of candidate donors; further, the post-fecal bacteria transplantation monitoring data is input into a transplantation strategy control model, and a transplantation adjustment strategy for adjusting the transplantation of the to-be-matched recipient in a next period is output; wherein the transplantation adjustment strategy comprises at least one of a donor switching strategy and a dose adjustment strategy. The technical solution provided in the embodiment is based on a dynamic feedback mechanism of real-time monitoring data after transplantation, continuously optimizes donor selection and dose adjustment through a transplantation strategy control model, solves the pain point that a static matching cannot adapt to dynamic changes after transplantation, in addition, integrates multi-modal data such as gene features, immune features and clinical features, breaks through the limitation of single microbiome data; comprehensive compatibility evaluation is realized through multi-modal data integration, and a treatment scheme is continuously optimized by means of a dynamic feedback mechanism, which significantly improves the precision and adaptability of FMT treatment.

[0154] The fecal bacteria transplantation donor-recipient dynamic matching device provided in the embodiment of the present application can execute the fecal bacteria transplantation donor-recipient dynamic matching method provided in any embodiment of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0155] Figure 5 A structural schematic diagram of an electronic device provided in the embodiment of the present application. Figure 5 The electronic device 50 shown is merely an example, and should not bring any limitation to the functions and use range of the embodiment of the present application. Figure 5As shown, the electronic device 50 is in the form of a general-purpose computing device. The components of the electronic device 50 can include, but are not limited to, one or more processors or processing units 501, a system memory 502, and a bus 503 that couples various system components including the system memory 502 to the processing unit 501. The bus 503 represents one or more of any of several bus structures, including a memory bus or memory controller, a peripheral bus, a graphics accelerator bus, and a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, Industrial Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus. The electronic device 50 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by the electronic device 50 and includes both volatile and non- volatile media, removable and non-removable media. The system memory 502 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 504 and / or cache memory 505. The electronic device 50 can further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, a storage system 506 can be provided for reading from and writing to a non-removable, non-volatile magnetic media (e.g., a "hard drive") Figure 5 (not shown), commonly referred to as a "hard disk drive"). Although not specifically shown, the electronic device 50 can further include other removable / non-removable, volatile / non-volatile computer system storage media including, but not limited to, a magnetic floppy disk drive (e.g., to read from and / or write to a removable magnetic floppy disk 527) and an optical disk drive (e.g., for reading from and / or writing to a removable optical disk 528 such as a CD-ROM or other optical media). The storage system 506 can be connected to the system bus 503 by a removable storage interface 509. The storage system 506 and the removable storage interface 509 can together also be considered to be a removable storage drive, wherein the removable storage interface 509 is a component that interfaces between the storage system 506 and the system bus 503. The storage system 506 and the storage interface 509 can be configured in a similar manner as the system memory 502 and the memory interface 503, respectively. Figure 5A disk drive, a floppy disk drive, and / or other mass storage device can be provided for in the system for reading from or writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk"), and to read from or write to a removable, nonvolatile optical disk such as a CD-ROM or other optical media. A disk drive or other mass storage device can be connected to the system bus 503 through an appropriate device adapter or interface. The RAM 502 can also include a high-speed RAM such as static RAM for caching data and lookup tables, etc. The RAM 502 can be used to store volatile data and / or execute programs in a high-speed RAM environment. An operating system 508, programs 507, and program data 509 can be stored in system memory 502, such as RAM, for execution by the processing unit(s) 501, according to embodiments of the application. The various program modules that can be stored in system memory 502 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each or a combination of which can include implementation of a network environment. The program modules 507, which have the program / utility 508, can be stored on the storage device (e.g., the disk drive 506), the RAM 502, or the like, and loaded into the system memory 502 using, for example, an optical disk drive, a floppy disk drive, etc., through an appropriate drive interface. The system memory 502, including the RAM 502, the ROM 505, and / or the cache memory (not shown), can be used as a general hard disk drive or a computer platform for processing and storage. Figure 5 Other hardware and / or software modules can be used in conjunction with the electronic device 50, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc. The processing unit(s) 501 can execute the various functions and / or methods described in the embodiments of the application by running programs from system memory 502. In particular, the processes described above with reference to the flow charts can be implemented as computer software programs. For example, the embodiments of the application include a computer program product which includes a computer program fixed in at least one computer readable medium that, when executed, performs methods in accordance with the embodiments of the application. In such an embodiment, the computer program can be downloaded from a network or installed from a storage system 506 via the I / O interface 511. When the computer program is executed by the processing unit(s) 501, it performs the functions described above as defined in the methods of the embodiments of the application.

[0156] The embodiment of the present application also provides a storage medium containing computer executable instructions, which are used for executing the above-mentioned fecal bacteria transplantation donor-recipient dynamic matching method when executed by a computer processor.

Claims

1. A method for dynamic matching of donor and recipient in fecal microbiota transplantation, characterized in that, The method comprises the following steps: In response to a donor-recipient matching request, obtaining post-fecal flora transplantation monitoring data of a recipient to be matched at a current preset sampling time; wherein the post-fecal flora transplantation monitoring data comprises recipient symptom score and flora colonization rate, and the post-fecal flora transplantation monitoring data is determined after a fecal flora transplantation operation is completed on the recipient to be matched based on the fecal flora of any one target donor, and the target donor is determined based on genetic feature information, immune feature information and clinical feature information of the recipient to be matched and a plurality of candidate donors. Inputting the post-fecal flora transplantation monitoring data into a transplantation strategy control model to output a transplantation adjustment strategy for adjusting the recipient to be matched in the next period; wherein the transplantation adjustment strategy comprises at least one of a donor switching strategy and a dose adjustment strategy.

2. The method of claim 1, wherein, The target donor is determined by the following method: Obtaining recipient sample feature data corresponding to the recipient to be matched and donor sample feature data corresponding to a plurality of candidate donors; wherein the recipient sample feature data comprises disease type and severity identification, recipient genetic feature information, recipient immune feature information and recipient clinical feature information; and the donor sample feature data comprises donor genetic feature information, donor immune feature information and donor clinical feature information. For each candidate donor, performing matching calculation on the recipient sample feature data and the donor sample feature data corresponding to the current candidate donor based on a pre-trained donor-recipient matching model to obtain microbial compatibility attributes and clinical adaptation attributes. Based on the disease type and severity identification of the recipient to be matched, the microbial compatibility attributes and the clinical adaptation attributes, determining the matching attributes between the recipient to be matched and the current candidate donor. Based on the matching attributes between the recipient to be matched and each candidate donor, determining a plurality of target donors adapted to the recipient to be matched.

3. The method of claim 1, wherein, The transplantation strategy control model comprises an intelligent agent decision network and a virtual environment network, and the intelligent agent decision network comprises an action value mapping layer and an action decision layer. The inputting of the post-fecal flora transplantation monitoring data into the transplantation strategy control model to output the transplantation adjustment strategy for adjusting the recipient to be matched in the next period comprises the following steps: Taking the post-fecal flora transplantation monitoring data as initial state information, and taking the initial state information as current state information; Inputting the current state information into the action value mapping layer to obtain value parameters corresponding to at least one preset execution action; wherein the at least one preset execution action comprises switching to each donor in the plurality of target donors and a random combination of at least one dose adjustment mode; Inputting each value parameter into the action decision layer to obtain a target value parameter satisfying a preset condition, and determining the preset execution action corresponding to the target value parameter as a current execution action. output an action reward value and next state information corresponding to the current execution action after the virtual environment network simulation executes the current execution action; wherein the action reward value is determined according to a clinical remission prediction rate, an adverse reaction prediction rate, and a bacterial diversity index prediction value; update the next state information as the current state information, and repeatedly execute the steps of determining the value parameter corresponding to the at least one preset execution action, determining the current execution action, and determining the action reward value and the next state information corresponding to the current execution action, until the action reward value obtained by repeating the steps a plurality of times meets a preset condition, and the current execution action is taken as a transplantation adjustment strategy for adjusting the recipient in a next time period.

4. The method of claim 2, wherein, The method further comprises: obtaining recipient sample data corresponding to the recipient to be matched and donor sample data corresponding to the plurality of candidate donors; wherein the recipient sample data comprises recipient fecal bacterial microbial gene data, recipient immune state data, and recipient clinical metadata; and the donor sample data comprises donor fecal bacterial microbial gene data, donor immune state data, and recipient clinical metadata; performing feature extraction processing on the recipient sample data to obtain recipient gene feature information, recipient immune feature information, and recipient clinical feature information; performing feature extraction processing on each of the donor sample data to obtain donor gene feature information, donor immune feature information, and donor clinical feature information corresponding to each of the candidate donors.

5. The method of claim 4, wherein, The recipient fecal bacterial microbial gene data comprises metagenomic sequencing data, species composition type of the recipient fecal sample, and relative abundance data of each species, and the recipient gene feature information comprises microbial species relative abundance feature, microbial co-occurrence network graph, and functional gene prediction feature information; performing feature extraction processing on the recipient sample data to obtain recipient gene feature information, comprising: performing feature vectorization processing on the species composition type of the recipient fecal sample and the relative abundance data of each species to obtain microbial species relative abundance feature; performing network graph construction processing on the species composition type of the recipient fecal sample and the relative abundance data of each species by a co-occurrence network analysis tool to obtain a microbial co-occurrence network graph; wherein the nodes in the microbial co-occurrence network graph represent different species, and the connection edges between the nodes represent significant correlations between species; performing gene function prediction analysis on the metagenomic sequencing data by a PICRUSt2 functional prediction tool to obtain functional gene prediction results, and performing feature vectorization processing on the functional gene prediction results to obtain functional gene prediction feature information.

6. The method of claim 4, wherein, performing feature extraction processing on the recipient sample data to obtain recipient clinical feature information, comprising: The donor-recipient matching model comprises a microbial compatibility evaluation model and a clinical adaptability evaluation model, and the recipient gene feature information comprises microbial co-occurrence network graph and functional gene prediction feature information.

7. The method of claim 2, wherein, The donor-recipient matching model comprises a microbial compatibility evaluation model and a clinical adaptability evaluation model, and the recipient gene feature information comprises microbial co-occurrence network graph and functional gene prediction feature information. The donor-recipient matching model comprises a microbial compatibility evaluation model and a clinical adaptability evaluation model, and the recipient gene feature information comprises microbial co-occurrence network graph and functional gene prediction feature information. The microbial compatibility attribute between the recipient to be processed and the current candidate donor is determined based on the recipient microbial species relative abundance feature, the recipient microbial co-occurrence network feature information in the recipient sample feature data, the donor microbial species relative abundance feature, the donor microbial co-occurrence network feature information in the donor sample feature data, and the microbial compatibility evaluation model. The clinical adaptability attribute between the recipient to be processed and the current candidate donor is determined based on the recipient functional gene prediction feature information, the recipient immune feature information, the recipient clinical feature information in the recipient sample feature data, the donor functional gene prediction feature information, the donor immune feature information, the donor clinical feature information in the donor sample feature data, and the clinical adaptability evaluation model.

8. The method of claim 7, wherein, The microbial compatibility attribute between the recipient to be processed and the current candidate donor is determined based on the recipient microbial species relative abundance feature, the recipient microbial co-occurrence network feature information in the recipient sample feature data, the donor microbial species relative abundance feature, the donor microbial co-occurrence network feature information in the donor sample feature data, and the microbial compatibility evaluation model. The relative abundance data corresponding to each microbial species and the evolutionary relationship coding information corresponding to each species in the recipient sample data are configured as node parameter information of the species corresponding to the recipient feature microbial co-occurrence network graph, so as to obtain the recipient feature microbial co-occurrence network graph. The relative abundance data corresponding to each microbial species and the evolutionary relationship coding information corresponding to each species in the donor sample data are configured as node parameter information of the species corresponding to the donor microbial co-occurrence network graph, so as to obtain the donor feature microbial co-occurrence network graph. The recipient feature microbial co-occurrence network graph and the donor feature microbial co-occurrence network graph are respectively input into the pre-trained graph neural network to obtain the recipient microbial feature embedding and the donor microbial feature embedding. The microbial compatibility attribute between the recipient to be processed and the current candidate donor is determined based on the similarity attribute between the recipient microbial feature embedding and the donor microbial feature embedding.

9. The method of claim 7, wherein, The clinical adaptation attribute between the to-be-processed recipient and the current candidate donor is determined based on the recipient function gene prediction feature information, the recipient immune feature information, the recipient clinical feature information in the recipient sample feature data, the donor function gene prediction feature information, the donor immune feature information, the donor clinical feature information in the donor sample feature data, and the clinical adaptability evaluation model, and includes: Based on the recipient function gene prediction feature information, the recipient immune feature information and the recipient clinical feature information in the recipient sample data, a recipient to-be-processed vector is constructed; Based on the donor function gene prediction feature information, the donor immune feature information and the donor clinical feature information in the donor sample data, a donor to-be-processed vector is constructed; The recipient to-be-processed vector and the donor to-be-processed vector are input into a pre-trained Transformer model to obtain a postoperative infection risk probability and a rejection reaction risk probability; Based on the postoperative infection risk probability and the rejection reaction risk probability, the clinical adaptation attribute between the to-be-processed recipient and the current candidate donor is determined.

10. The method of claim 2, wherein, The matching attribute between the to-be-matched recipient and the current candidate donor is determined based on the disease type information and the severity label information of the to-be-matched recipient, the microorganism compatibility attribute and the clinical adaptation attribute, and includes: Based on the disease type information and the severity label information of the to-be-matched recipient, a microorganism compatibility weight and a clinical adaptation weight are determined; The microorganism compatibility attribute and the clinical adaptation attribute are weighted and operated based on the microorganism compatibility weight and the clinical adaptation weight, and the matching attribute between the to-be-matched recipient and the current candidate donor.

11. The method of claim 10, wherein, The microorganism compatibility weight and the clinical adaptation weight are determined based on the disease type information and the severity label information of the to-be-matched recipient, and include: If the to-be-matched recipient is a primary patient based on the disease type information and the severity label information of the to-be-matched recipient, the microorganism compatibility weight is determined from a first preset numerical range, and the clinical adaptation weight is determined from a second preset numerical range; wherein the numerical values in the first preset numerical range are all greater than the numerical values in the second preset numerical range; If the to-be-matched recipient is a refractory patient based on the disease type information and the severity label information of the to-be-matched recipient, the microorganism compatibility weight is determined from a third preset numerical range, and the clinical adaptation weight is determined from a fourth preset numerical range; wherein the numerical values in the third preset numerical range are all smaller than the numerical values in the fourth preset numerical range.

12. A dynamic matching device for fecal microbiota transplantation donors and recipients, characterized in that, The device includes: The monitoring data acquisition module is configured to acquire post-fecal bacteria transplantation monitoring data of the recipient to be matched at a current preset sampling time in response to a request for matching a donor and a recipient, wherein the post-fecal bacteria transplantation monitoring data comprises a symptom score of the recipient and a colonization rate of a bacterial community, and the post-fecal bacteria transplantation monitoring data is determined based on a fecal bacterial community of any one target donor after a fecal bacteria transplantation operation is completed on the recipient to be matched, and the target donor is determined based on genetic feature information, immune feature information, and clinical feature information of the recipient to be matched and a plurality of candidate donors. The adjustment strategy determination module is configured to input the post-fecal bacteria transplantation monitoring data into a transplantation strategy control model, and output a transplantation adjustment strategy for adjusting the recipient to be matched in a next period, wherein the transplantation adjustment strategy comprises at least one of a donor switching strategy and a dose adjustment strategy.

13. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the fecal bacteria transplantation donor-recipient dynamic matching method of any one of claims 1-11.

14. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for enabling the processor to execute the fecal bacteria transplantation donor-recipient dynamic matching method of any one of claims 1-11 when executed.

Images