Recombinant risk assessment intelligent analysis platform for wild bird influenza virus

Abstract

The application provides an intelligent analysis platform for risk assessment of wild bird flu virus recombination, relates to the technical field of influenza virus monitoring and risk assessment, and comprises a multi-source monitoring data acquisition module, a data preprocessing module, a virus sequence intelligent analysis module, a recombination risk assessment module, a spatial situation analysis and visualization module and a report generation and early warning module connected in sequence through an API interface. The intelligent analysis platform for risk assessment of wild bird flu virus recombination can realize multi-source monitoring data acquisition, virus sequence intelligent analysis, recombination risk quantitative assessment and spatial visualization early warning, and provides whole-process technical support for influenza virus recombination prevention and control.

Description

Technical Field

[0001] This invention relates to the field of influenza virus monitoring and risk assessment technology, specifically to an intelligent analysis platform for risk assessment of wild avian influenza virus recombination. Background Technology

[0002] For poultry farming, veterinary departments across the country regularly collect throat swabs and fecal samples from poultry farms for rapid laboratory testing to confirm the presence of influenza virus strains. For wild birds, research institutions frequently conduct migratory bird tracking and on-site sampling, sending the collected virus samples to virus research centers for high-throughput sequencing and genome assembly. In terms of data analysis, many laboratories have established influenza virus sequence comparison and subtype identification processes, using public databases for comparison to determine the subtype of a sample and its possible phylogenetic relationships. Some studies have attempted to apply statistical models or sequence feature-based machine learning methods to conduct source tracing analysis of genetic distances and evolutionary paths between different subtypes. Furthermore, some large-scale research projects and international collaborations (such as the "Wildlife Influenza Surveillance Program") combine migratory bird migration routes, climate and environmental data, and poultry farming distribution information to conduct spatial trend analysis of influenza transmission using geographic information systems, and publish regular monitoring reports and epidemiological assessments. Although these technologies have played an important role in monitoring existing influenza subtypes in wild animals and poultry, they are mostly focused on the detection and monitoring of known strains, and a specialized assessment system for gene recombination events has not yet been established.

[0003] The above monitoring and analysis process still has shortcomings overall:

[0004] Insufficient monitoring coverage: Wildlife (especially migratory birds) sampling sites are scattered and have a wide migration range, making it difficult for the existing monitoring network to fully cover them, resulting in potential recombinant sources not being included in the analysis;

[0005] Low analysis efficiency: Traditional laboratory high-throughput sequencing and manual comparison processes are time-consuming and cannot meet the needs of real-time monitoring;

[0006] Recombination detection deficiencies: There is a lack of efficient and automated algorithms for detecting recessive recombination fragments, which can easily miss key recombination signals;

[0007] Lack of assessment system: Current technology mainly relies on monitoring known strains and has not formed a dedicated assessment system for viral recombination events, making it impossible to quantify the recombination risk level.

[0008] To address the issues raised above, we have developed an intelligent analysis platform for assessing the risk of wild avian influenza virus recombination. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides an intelligent analysis platform for assessing the recombination risk of avian influenza viruses. The technical problems this invention aims to solve are: incomplete monitoring coverage, low analysis efficiency, weak detection capability for latent recombination, and the lack of a recombination risk assessment system.

[0010] To achieve the above objectives, the present invention provides the following technical solution:

[0011] A smart analysis platform for risk assessment of wild avian influenza virus recombination includes:

[0012] The multi-source monitoring data acquisition module, data preprocessing module, virus sequence intelligent analysis module, recombination risk assessment module, spatial situation analysis and visualization module, and report generation and early warning module are connected sequentially through the API interface.

[0013] Among them, the multi-source monitoring data acquisition module is used to collect data in multiple dimensions, covering monitoring scenarios for wild animals and poultry;

[0014] The data preprocessing module is used to clean and standardize the collected data, eliminate noise interference, and provide high-quality data for subsequent analysis.

[0015] The intelligent virus sequence analysis module is used for efficient sequencing analysis, detection of recessive recombination fragments, and subtype identification.

[0016] The restructuring risk assessment module is used to establish a restructuring risk assessment system and quantify the risk level by comprehensively considering multiple factors.

[0017] The spatial situation analysis and visualization module is used to intuitively display the spatial characteristics of virus distribution and recombination risk, and to enable dynamic tracking;

[0018] The report generation and early warning module is used to automatically generate assessment reports and provide real-time early warnings for high-risk events.

[0019] Preferably, the multi-source monitoring data acquisition module includes a mobile sampling terminal, a fixed monitoring node, and an environmental data interface;

[0020] The mobile sampling terminal integrates GPS positioning, sample information entry and sequencing data preprocessing functions, and is deployed at key migratory bird stopover sites.

[0021] The fixed monitoring nodes are deployed in poultry farms to automatically collect sample sequencing data and stocking density data;

[0022] The environmental data interface is used to connect with meteorological departments and GIS systems to obtain environmental data.

[0023] Preferably, the intelligent virus sequence analysis module includes a high-throughput sequencing rapid parsing submodule, a recessive recombination fragment detection submodule, and a sequence alignment and subtype identification submodule;

[0024] The high-throughput sequencing rapid parsing submodule, employing a parallel computing framework and based on the SOAPdenovo2 optimization algorithm, achieves rapid genome assembly and further includes:

[0025] Data preprocessing: The raw sequencing data undergoes quality filtering to remove low-quality reads and adapter sequences. Base quality scores are calculated using the FastQC tool. Calculation formula:

[0026] ,in Set a threshold for the probability of base errors. ≥30 retain high-quality data.

[0027] Constructing a K-mer library: The filtered data is segmented according to different K-mer lengths to generate a K-mer frequency distribution map. The DeBruijn graph algorithm is used to construct the overlap relationships between K-mers, where nodes represent K-mers and edges represent overlapping connections between K-mers.

[0028] Genome assembly: Based on the SOAPdenovo2 optimization algorithm, the DeBruijn graph is used for path search to assemble K-mers into contigs. The optimized assembly process introduces an iterative correction mechanism to calculate the coverage between contigs.

[0029] The formula for the SOAPdenovo2 optimization algorithm is as follows:

[0030] ,when When the value is ≥0.8, the data is merged to generate a Scaffold.

[0031] Parallel computing acceleration: Employing the MPI parallel computing framework, computationally intensive tasks such as K-mer partitioning, DeBruijn graph construction, and path search are distributed to multiple nodes for parallel processing, theoretically achieving a speedup of [missing information]. ,in For single-node processing time, for Parallel processing time of each node;

[0032] The latent recombination fragment detection submodule constructs a Transformer-based deep learning model, uses known recombinant virus sequences as a training set, learns the sequence features of latent recombination fragments, and automatically identifies latent recombination regions in samples.

[0033] The formula for the deep learning model is as follows:

[0034] Let the number of correctly detected recessive recombination regions be . The actual total number of hidden recombination regions is The formula for calculating the detection accuracy is:

[0035] This formula is used to calculate the accuracy of the test results;

[0036] in, This indicates the number of samples with correct test results. This represents the total number of samples tested. The accuracy rate is calculated by dividing the number of correct samples by the total number of samples and then multiplying by 100%. The formula requires that the accuracy rate reach or exceed 92% to ensure that the test results have high reliability.

[0037] The sequence alignment and subtype identification submodule is used to interface with the NCBI influenza virus database, and adopts an improved BLAST algorithm to achieve rapid subtype identification and output a phylogenetic matrix between the sample and known strains.

[0038] The formula for the BLAST algorithm is as follows:

[0039] This formula is used to calculate the proportion of time reduction in comparison, to achieve rapid subtype identification, and to output a phylogenetic matrix between the sample and known strains;

[0040] In the formula, the time reduction ratio represents the percentage reduction in time after optimization or the adoption of a new method compared to the original time.

[0041] : Represents the original time required to complete a task, experiment, etc., before optimization or modification;

[0042] : Represents the time required to complete the same task or experiment after optimization or modification;

[0043] : Calculate the proportion of the changed time to the original time;

[0044] The percentage reduction in time can be obtained by subtracting the percentage of time that has been changed from 1.

[0045] Convert the time reduction percentage to a percentage form to more intuitively reflect the degree of time reduction.

[0046] Preferably, the recombination risk assessment module includes a genetic distance calculation submodule, an evolutionary path tracing submodule, and a risk level assessment submodule;

[0047] The genetic distance calculation submodule, based on the k-mer frequency improved Jukes-Cantor model, quickly calculates the genetic distance between wild animal and poultry virus strains;

[0048] The evolutionary path tracing submodule uses a Bayesian evolutionary analysis model, combined with sample collection time and geographical information, to construct a virus evolutionary tree and trace possible recombination sources.

[0049] The risk level assessment submodule establishes a three-dimensional assessment index system, calculates weights using the analytic hierarchy process, and outputs three levels of risk: low risk, medium risk, and high risk.

[0050] Preferably, the spatial situation analysis and visualization module is developed based on ArcGIS Engine, overlaying migratory bird migration routes, poultry farming areas, sampling points and risk level layers, and dynamically displaying risk trends through a timeline.

[0051] Preferably, when the risk level is high, the report generation and early warning module pushes early warning information via SMS and email, and simultaneously outputs prevention and control suggestions.

[0052] This invention provides an intelligent analysis platform for risk assessment of wild animal-avian influenza virus recombination. It has the following beneficial effects:

[0053] This invention uses a multi-source acquisition mode of mobile sampling terminals + fixed monitoring nodes + environmental data interfaces to cover migratory bird routes and poultry farming areas, thus solving the problem of insufficient monitoring networks.

[0054] This invention employs MPI parallel computing, SOAPdenovo2 optimization algorithm, and improved BLAST algorithm to shorten genome assembly time by more than 60% and subtype identification time by ≥66.7%, meeting the needs of real-time monitoring.

[0055] This invention achieves a latent recombination fragment detection accuracy of ≥92% based on the Transformer deep learning model, avoiding the omission of key recombination signals;

[0056] This invention establishes a three-dimensional assessment index system and quantifies risk levels through AHP weight calculation, providing data support for prevention and control decisions.

[0057] This invention provides real-time early warning information and prevention and control suggestions for high-risk events, facilitating rapid response and reducing the risk of viral recombination and transmission. Attached Figure Description

[0058] Figure 1This is a schematic diagram of the system architecture for implementing the invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] like Figure 1 As shown, this embodiment of the invention provides an intelligent analysis platform for assessing the risk of wild avian influenza virus recombination, including:

[0061] The multi-source monitoring data acquisition module, data preprocessing module, virus sequence intelligent analysis module, recombination risk assessment module, spatial situation analysis and visualization module, and report generation and early warning module are connected sequentially through the API interface.

[0062] Among them, the multi-source monitoring data acquisition module is used to collect data in multiple dimensions, covering monitoring scenarios for wild animals and poultry;

[0063] The multi-source monitoring data acquisition module includes a mobile sampling terminal, a fixed monitoring node, and an environmental data interface;

[0064] The mobile sampling terminal integrates GPS positioning, sample information entry and sequencing data preprocessing functions, and is deployed at key migratory bird stopover sites.

[0065] The fixed monitoring nodes are deployed in poultry farms to automatically collect sample sequencing data and stocking density data;

[0066] The environmental data interface is used to connect with meteorological departments (temperature, humidity, precipitation) and GIS systems (migratory bird routes, poultry farming area distribution) to obtain environmental data.

[0067] The data preprocessing module is used to clean and standardize the collected data, eliminate noise interference, and provide high-quality data for subsequent analysis.

[0068] Sequencing data quality control: FastQC was used for quality detection, and low-quality reads (Q value < 20) and adapter sequences were removed by trimmomatic analysis.

[0069] Unstructured data standardization: Convert sample information (species, sampling points) and environmental data into a unified format (such as JSON) and associate them with GPS coordinates and timestamps;

[0070] Data deduplication: Based on sample ID and sequencing fragment characteristics, duplicate data is removed to avoid redundant calculations.

[0071] The intelligent virus sequence analysis module is used for efficient sequencing analysis, detection of recessive recombination fragments, and subtype identification.

[0072] The intelligent virus sequence analysis module includes a high-throughput sequencing rapid parsing submodule, a recessive recombination fragment detection submodule, and a sequence alignment and subtype identification submodule;

[0073] The high-throughput sequencing rapid parsing submodule, employing a parallel computing framework and based on the SOAPdenovo2 optimization algorithm, achieves rapid genome assembly and further includes:

[0074] Data preprocessing: The raw sequencing data undergoes quality filtering to remove low-quality reads and adapter sequences. Base quality scores are calculated using the FastQC tool. Calculation formula:

[0075] ,in Set a threshold for the probability of base errors. ≥30 retain high-quality data.

[0076] Constructing a K-mer library: The filtered data is segmented according to different K-mer lengths (e.g., K=21, 31, 41) to generate a K-mer frequency distribution map. The DeBruijn graph algorithm is used to construct the overlap relationships between K-mers, where nodes represent K-mers and edges represent overlapping connections between K-mers.

[0077] Genome assembly: Based on the SOAPdenovo2 optimization algorithm, path search is performed using the DeBruijn graph to assemble K-mers into contigs. The optimized assembly process incorporates an iterative correction mechanism, calculating the coverage between contigs.

[0078] ,when When the value is ≥0.8, the components are merged to generate Scaffold.

[0079] Parallel computing acceleration: Employing the MPI parallel computing framework, computationally intensive tasks such as K-mer partitioning, DeBruijn graph construction, and path search are distributed to multiple nodes for parallel processing, theoretically achieving a speedup of [missing information]. ,in For single-node processing time, for The time required for parallel processing by each node.

[0080] The latent recombination fragment detection submodule constructs a Transformer-based deep learning model, uses known recombinant virus sequences as a training set, learns the sequence features of latent recombination fragments, and automatically identifies latent recombination regions in samples.

[0081] The formula for the deep learning model is as follows:

[0082] Let the number of correctly detected recessive recombination regions be . The actual total number of hidden recombination regions is The formula for calculating the detection accuracy is:

[0083] This formula is used to calculate the accuracy of the test results;

[0084] in, This indicates the number of samples with correct test results. This represents the total number of samples tested. The accuracy rate is calculated by dividing the number of correct samples by the total number of samples and then multiplying by 100%. The formula requires that this accuracy rate reach or exceed 92% to ensure that the test results are highly reliable.

[0085] The sequence alignment and subtype identification submodule is used to interface with the NCBI influenza virus database, adopts an improved BLAST algorithm to achieve rapid subtype identification, and outputs a phylogenetic matrix between the sample and known strains.

[0086] The formula for the BLAST algorithm is:

[0087] This formula is used to calculate the proportion of time reduction in comparison, to achieve rapid subtype identification, and to output a phylogenetic matrix between the sample and known strains;

[0088] In the formula, the time reduction ratio represents the percentage reduction in time compared to the original time after optimization or the adoption of a new method, expressed as a percentage.

[0089] This represents the original time required to complete a task, experiment, etc., before optimization or modification.

[0090] This represents the time required to complete the same task or experiment after optimization or modification.

[0091] : Calculate the proportion of the changed time to the original time.

[0092] The percentage of time shortened is obtained by subtracting the percentage of time changed from 1.

[0093] Convert the time reduction percentage to a percentage form to more intuitively reflect the degree of time reduction.

[0094] The formula sets a criterion, requiring that the reduction in comparison time must be greater than or equal to... (Approximately 66.7%), meaning that the time reduction must reach or exceed this standard to meet the requirements.

[0095] The restructuring risk assessment module is used to establish a restructuring risk assessment system and quantify the risk level by comprehensively considering multiple factors.

[0096] The recombination risk assessment module includes a genetic distance calculation submodule, an evolutionary path tracing submodule, and a risk level assessment submodule;

[0097] The genetic distance calculation submodule, based on the k-mer frequency improved Jukes-Cantor model, quickly calculates the genetic distance between wild animal and poultry virus strains. If the distance is <0.05 (i.e. sequence similarity >95%), it is marked as "potential recombination association".

[0098] Data preprocessing: The nucleic acid sequences of wild animal and poultry virus strains were standardized to remove redundant fragments and low-quality bases.

[0099] k-mer extraction: The viral sequence is divided into fixed-length k-mers (e.g., k=21), and the frequency of each k-mer in the sequence is counted.

[0100] Frequency matrix construction: Construct a frequency similarity matrix between different viral strains based on k-mer frequencies.

[0101] Genetic distance calculation: Based on the improved Jukes-Cantor model, using the formula Genetic distance calculation is performed. The proportion of k-mer frequency differences between sequences .

[0102] Risk marker: If the calculated genetic distance is <0.05 (corresponding sequence similarity >95%), it is marked as "potential recombination association".

[0103] The evolutionary path tracing submodule uses a Bayesian evolutionary analysis model (BEAST) and combines sample collection time and geographical information to construct a virus evolutionary tree and trace possible recombination sources.

[0104] Data integration: Collect nucleic acid sequences, collection time, geographical coordinates and host information of virus samples to construct a comprehensive dataset.

[0105] Parameter settings: Set prior parameters in the BEAST model, including molecular clock models (such as strict molecular clock or relaxed molecular clock) and tree prior models (such as Yule tree or birth-death tree).

[0106] MCMC execution: Iterative calculations are performed using the Markov Chain Monte Carlo (MCMC) algorithm, running more than 10 million times to ensure that the effective sample size (ESS) is greater than 200.

[0107] Evolutionary tree construction: The convergence of MCMC is checked using Tracer software, and the posterior tree is integrated using TreeAnnotator to generate the phylogenetic tree with the highest confidence.

[0108] Recombination origin tracing: By combining geographic information systems (GIS), the spatiotemporal information of key nodes on the evolutionary tree is marked to trace the time, location and host conversion path of viral recombination events.

[0109] The risk level assessment submodule establishes a three-dimensional assessment index system of "virus characteristics - geographical association - environmental adaptation", calculates the weights through the analytic hierarchy process (AHP), and outputs three levels of risk: low risk, medium risk, and high risk.

[0110] Evaluation Dimensions Indicator Examples Weight Virus characteristics Number of recessive recombinant fragments, subtype pathogenicity 0.4 Geographical association Are the sampling points located in the overlapping area of ​​migratory bird routes and breeding areas? 0.3 Environment adaptation Temperature (15-25℃ is the suitable range for virus survival), humidity (>60%). 0.3

[0111] The spatial situation analysis and visualization module is used to intuitively display the spatial characteristics of virus distribution and recombination risk, and to enable dynamic tracking;

[0112] The spatial situation analysis and visualization module is developed based on ArcGIS Engine. It overlays migratory bird migration routes, poultry farming areas, sampling points, and risk level layers (low risk: green, medium risk: yellow, high risk: red) and dynamically displays risk trends through a timeline.

[0113] The report generation and early warning module is used to automatically generate assessment reports and provide real-time early warnings for high-risk events.

[0114] The report generation and early warning module automatically integrates sample information, sequence analysis results, risk level and spatial situation map to generate a monitoring report that conforms to epidemiological standards. When the risk level is high, it pushes early warning information via SMS and email and simultaneously outputs prevention and control suggestions.

[0115] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart analysis platform for risk assessment of wild avian influenza virus recombination, characterized in that: include: The multi-source monitoring data acquisition module, data preprocessing module, virus sequence intelligent analysis module, recombination risk assessment module, spatial situation analysis and visualization module, and report generation and early warning module are connected sequentially through the API interface. Among them, the multi-source monitoring data acquisition module is used to collect data in multiple dimensions, covering monitoring scenarios for wild animals and poultry; The data preprocessing module is used to clean and standardize the collected data, eliminate noise interference, and provide high-quality data for subsequent analysis. The intelligent virus sequence analysis module is used for efficient sequencing analysis, detection of recessive recombination fragments, and subtype identification. The restructuring risk assessment module is used to establish a restructuring risk assessment system and quantify the risk level by comprehensively considering multiple factors. The spatial situation analysis and visualization module is used to intuitively display the spatial characteristics of virus distribution and recombination risk, and to enable dynamic tracking; The report generation and early warning module is used to automatically generate assessment reports and provide real-time early warnings for high-risk events.

2. The intelligent analysis platform for assessing the recombinant risk of avian influenza virus according to claim 1, characterized in that, The multi-source monitoring data acquisition module includes a mobile sampling terminal, a fixed monitoring node, and an environmental data interface; The mobile sampling terminal integrates GPS positioning, sample information entry and sequencing data preprocessing functions, and is deployed at key migratory bird stopover sites. The fixed monitoring nodes are deployed in poultry farms to automatically collect sample sequencing data and stocking density data; The environmental data interface is used to connect with meteorological departments and GIS systems to obtain environmental data.

3. The intelligent analysis platform for assessing the recombinant risk of avian influenza virus according to claim 1, characterized in that, The intelligent virus sequence analysis module includes a high-throughput sequencing rapid parsing submodule, a recessive recombination fragment detection submodule, and a sequence alignment and subtype identification submodule; The high-throughput sequencing rapid parsing submodule, employing a parallel computing framework and based on the SOAPdenovo2 optimization algorithm, achieves rapid genome assembly and further includes: Data preprocessing: The raw sequencing data undergoes quality filtering to remove low-quality reads and adapter sequences. Base quality scores are calculated using the FastQC tool. Calculation formula: ,in Set a threshold for the probability of base errors. ≥30 retain high-quality data.

4. Constructing a K-mer library: The filtered data is segmented according to different K-mer lengths to generate a K-mer frequency distribution map. The DeBruijn graph algorithm is used to construct the overlap relationship between K-mers, where nodes represent K-mers and edges represent overlapping connections between K-mers; Genome assembly: Based on the SOAPdenovo2 optimization algorithm, the DeBruijn graph is used for path search to assemble K-mers into contigs. The optimized assembly process introduces an iterative correction mechanism to calculate the coverage between contigs. The formula for the SOAPdenovo2 optimization algorithm is as follows: ,when When the value is ≥0.8, the data is merged to generate a Scaffold. Parallel computing acceleration: Employing the MPI parallel computing framework, computationally intensive tasks such as K-mer partitioning, DeBruijn graph construction, and path search are distributed to multiple nodes for parallel processing, theoretically achieving a speedup of [missing information]. ,in For single-node processing time, for Parallel processing time of each node; The latent recombination fragment detection submodule constructs a Transformer-based deep learning model, uses known recombinant virus sequences as a training set, learns the sequence features of latent recombination fragments, and automatically identifies latent recombination regions in samples. The formula for the deep learning model is as follows: Let the number of correctly detected recessive recombination regions be . The actual total number of hidden recombination regions is The formula for calculating the detection accuracy is: This formula is used to calculate the accuracy of the test results; in, This indicates the number of samples with correct test results. This represents the total number of samples tested. The accuracy rate is calculated by dividing the number of correct samples by the total number of samples and then multiplying by 100%. The formula requires that the accuracy rate reach or exceed 92% to ensure that the test results have high reliability. The sequence alignment and subtype identification submodule is used to interface with the NCBI influenza virus database, and adopts an improved BLAST algorithm to achieve rapid subtype identification and output a phylogenetic matrix between the sample and known strains. The formula for the BLAST algorithm is as follows: This formula is used to calculate the proportion of time reduction in comparison, to achieve rapid subtype identification, and to output a phylogenetic matrix between the sample and known strains; In the formula, the time reduction ratio represents the percentage reduction in time after optimization or the adoption of a new method compared to the original time. : Represents the original time required to complete a task, experiment, etc., before optimization or modification; : Represents the time required to complete the same task or experiment after optimization or modification; : Calculate the proportion of the changed time to the original time; The percentage reduction in time can be obtained by subtracting the percentage of time that has been changed from 1. Convert the time reduction percentage to a percentage form to more intuitively reflect the degree of time reduction.

5. The intelligent analysis platform for assessing the recombinant risk of avian influenza virus according to claim 1, characterized in that, The recombination risk assessment module includes a genetic distance calculation submodule, an evolutionary path tracing submodule, and a risk level assessment submodule; The genetic distance calculation submodule, based on the k-mer frequency improved Jukes-Cantor model, quickly calculates the genetic distance between wild animal and poultry virus strains; The evolutionary path tracing submodule uses a Bayesian evolutionary analysis model, combined with sample collection time and geographical information, to construct a virus evolutionary tree and trace possible recombination sources. The risk level assessment submodule establishes a three-dimensional assessment index system, calculates weights using the analytic hierarchy process, and outputs three levels of risk: low risk, medium risk, and high risk.

6. The intelligent analysis platform for assessing the recombinant risk of avian influenza virus according to claim 1, characterized in that, The spatial situation analysis and visualization module is developed based on ArcGIS Engine, overlaying migratory bird migration routes, poultry farming areas, sampling points, and risk level layers, and dynamically displaying risk trends through a timeline.

7. The intelligent analysis platform for assessing the recombinant risk of avian influenza virus according to claim 1, characterized in that, When the risk level is high, the report generation and early warning module will push early warning information via SMS and email, and simultaneously output prevention and control suggestions.

Images