Multi-omics data optimization method and system for intestinal flora matching and medium

Abstract

The invention relates to the technical field of biological information mining, in particular to a multi-omics data optimization method and system for intestinal flora matching and a medium. The method comprises the following steps: acquiring historical intestinal tract multi-omics data and integrating the historical intestinal tract multi-omics data to obtain host intestinal tract multi-omics data; performing multi-omics data fusion based on the host intestinal multi-omics data to obtain microbial flora fusion data; performing pattern classification according to the host intestinal tract multi-omics data to obtain host intestinal tract physiological pattern data; carrying out physiological pattern influence factor calculation on the microbial flora fusion data to obtain host intestinal tract physiological pattern influence factors; and performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data according to the host intestinal physiological pattern influence factor, thereby obtaining intestinal flora matching data. According to the method, the accuracy of flora matching prediction can be improved.

Description

Technical Field

[0001] The present invention relates to the field of bioinformation mining technology, and in particular to a multi-omics data optimization method, system and medium for intestinal flora matching. Background Art

[0002] The composition of the gut microbiota is closely linked to host health. Recent studies have shown that the diversity of microbial communities and their interactions with host physiological characteristics significantly influence the development and progression of various diseases. Therefore, profiling the gut microbiota and exploring its relationship with host physiological data has become a hot topic in modern medical research. The diversity and complexity of the gut microbiota make it difficult for traditional single-omics analysis methods to fully reveal its biological significance. Therefore, integrating multi-omics data for comprehensive analysis is becoming an inevitable trend. Currently, research on the gut microbiota encompasses multiple technical fields, including but not limited to microbiomics, metabolomics, transcriptomics, and genomics. These technologies provide powerful tools for quantitative and qualitative analysis of the gut microbiome. However, despite the increasing application of various omics technologies, data integration and optimization remain challenging in practical research. Effectively integrating data from different omics sources and improving the predictive power of profiling models are both pressing challenges. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a multi-omics data optimization method, system and medium for intestinal flora matching to solve at least one of the above technical problems.

[0004] To achieve the above objectives, a multi-omics data optimization method for intestinal flora typing is provided, comprising the following steps:

[0005] Step S1: Acquire historical intestinal multi-omics data, and classify intestinal microbial flora based on the historical intestinal multi-omics data, thereby obtaining microbial flora multi-omics data; integrate the diversity of host intestinal flora based on the microbial flora multi-omics data, thereby obtaining host intestinal multi-omics data;

[0006] Step S2: performing microbial flora intestinal distribution statistics based on the host intestinal multi-omics data to obtain microbial flora intestinal distribution data, and performing multi-omics data fusion on the microbial flora multi-omics data according to the microbial flora intestinal distribution data to obtain microbial flora fusion data;

[0007] Step S3: extracting host intestinal physiological characteristics based on the host intestinal multi-omics data to obtain host intestinal physiological data, and classifying the host intestinal physiological pattern based on the host intestinal physiological data to obtain host intestinal physiological pattern data;

[0008] Step S4: performing high-dimensional omics semantic embedding on the microbial flora fusion data to obtain microbial flora dimensionality reduction data; calculating the physiological pattern influencing factors based on the host intestinal physiological pattern data and the microbial flora dimensionality reduction data to obtain the host intestinal physiological pattern influencing factors;

[0009] Step S5: performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data according to the influencing factors of the host intestinal physiological pattern, thereby obtaining intestinal flora matching data.

[0010] By collecting and analyzing historical intestinal multi-omics data, the present invention can establish a comprehensive microbial community classification system, which helps to reveal the diversity and functions of different microorganisms in the host intestine, thereby providing a solid data foundation for subsequent analysis. Statistics on the distribution of microbial flora can help understand the abundance of different microorganisms in the intestine and their ecological niches, which helps to identify specific microbial groups related to the host's health status and provides a basis for exploring their potential biomarkers. Fusion of multi-omics data of microbial flora can integrate information at different levels and reveal the association between microorganisms and host physiological characteristics. By extracting the host's intestinal physiological characteristics, important information about the host's status can be provided for subsequent analysis. These data will help researchers identify the relationship between intestinal flora and host physiological status, and further reveal the mechanism by which intestinal microorganisms affect host health. High-dimensional semantic embedding and dimensionality reduction processing of microbial flora data can effectively reduce the complexity of the data, making subsequent pattern recognition and analysis more efficient. This method helps to improve the interpretability of the data and simplifies the construction of subsequent matching models while ensuring the integrity of the information. By comprehensively analyzing the host's physiological patterns and microbial flora data, key factors affecting host health can be identified. This process not only helps with personalized matching of intestinal flora, but also simplifies the construction of matching models and provides theoretical support, which helps improve the predictive ability of flora matching models. Overall, these steps provide a comprehensive perspective on the relationship between intestinal flora and host health status by integrating multi-omics data and analyzing host physiological characteristics. This multi-dimensional research approach can reveal more complex biological mechanisms and promote a deeper understanding of intestinal microorganisms.

[0011] Optionally, step S1 specifically includes:

[0012] Step S11: acquiring historical intestinal multi-omics data and performing data cleaning on the historical intestinal multi-omics data to obtain intestinal multi-omics data to be classified;

[0013] Step S12: extracting gene sequencing data and host physiological coefficients from the intestinal multi-omics data to be classified, thereby obtaining intestinal gene sequencing data and host physiological coefficient data;

[0014] Step S13: classifying the microbial flora according to the intestinal gene sequencing data, thereby obtaining microbial flora classification data;

[0015] Step S14: performing microbial flora multi-omics classification integration on the intestinal multi-omics data to be classified according to the microbial flora classification data, thereby obtaining microbial flora multi-omics data;

[0016] Step S15: Integrate the diversity of the host intestinal flora according to the host physiological coefficient data and the microbial flora multi-omics data, thereby obtaining the host intestinal multi-omics data.

[0017] The present invention can provide basic data support for research by acquiring historical intestinal multi-omics data. The data cleaning process ensures data quality, removes noise and redundant information, and makes subsequent analysis more reliable and accurate, which lays a solid foundation for subsequent classification and analysis. By extracting gene sequencing data and host physiological coefficients from the intestinal multi-omics data to be classified, key biological information can be obtained, which not only helps to analyze the relationship between microorganisms and hosts, but also provides the necessary biological basis for understanding the function of the intestinal ecosystem. Classification of microbial flora based on intestinal gene sequencing data can identify and quantify the microbial composition in the intestine. This classification result helps to reveal the diversity and function of different microbial species, thereby providing an important basis for subsequent functional analysis and health research. By integrating microbial flora classification data, more comprehensive microbial flora multi-omics data can be formed. This integration can provide multi-level information to help researchers understand the role of microorganisms in the intestinal ecology and their potential impact on host health. By integrating host physiological coefficient data with microbial flora multi-omics data, we can deeply analyze the complex interactions between the host and intestinal microorganisms. This integration can not only reveal the relationship between the host's health status and microbial diversity, but also help develop applications in fields such as personalized medicine and precision nutrition.

[0018] Optionally, step S13 is specifically as follows:

[0019] Step S131: performing 16S rRNA sequence splicing on the intestinal gene sequencing data to obtain gene sequence splicing data;

[0020] Step S132: removing redundant sequences based on the gene sequence splicing data, thereby obtaining amplicon sequence variant data;

[0021] Step S133: obtaining microbial gene sequence reference data, and performing gene sequence similarity calculation on the microbial gene sequence reference data and amplicon sequence variant data, thereby obtaining gene sequence similarity data;

[0022] Step S134: performing microbial species classification on the amplicon sequence variant data according to the gene sequence similarity data, thereby obtaining microbial species classification data;

[0023] Step S135: performing microbial species relative abundance calculation on the microbial species classification data, thereby obtaining microbial species relative abundance data;

[0024] Step S136: dividing the microbial species classification data into microbial flora according to the microbial species relative abundance data, thereby obtaining microbial flora classification data.

[0025] The present invention can generate a complete gene sequence by splicing the 16S rRNA sequence, which helps to improve the accuracy and comprehensiveness of subsequent analysis and ensure that the information obtained is more reliable. After removing redundant sequences, the amplicon sequence variant data obtained reduces data noise, improves the quality of the data, and makes subsequent analysis more focused on actual microbial diversity. By calculating the similarity between the amplicon sequence variant data and the microbial gene sequence reference data, sequences similar to known microorganisms can be identified and classified, thereby providing a basis for species classification. Classifying microbial species based on similarity data can clarify the types of microorganisms present in the sample, enhance the understanding of the composition of intestinal microorganisms, and provide basic data for health research. Calculating the relative abundance of microbial species classification data can quantify the proportion of each microorganism in the sample, thereby analyzing its importance and role in intestinal ecology. By dividing the bacterial community based on the relative abundance data of microbial species, specific microbial community structures can be identified, and their relationship with host health can be further explored, providing guidance for potential intervention strategies.

[0026] Optionally, step S15 is specifically as follows:

[0027] Based on the host physiological coefficient data and the microbial flora multi-omics data, the host flora multi-omics association is performed to obtain the host intestinal flora multi-omics data;

[0028] Perform Shannon Index diversity analysis on the multi-omics data of the host intestinal flora to obtain the Shannon Index data of the host intestinal flora;

[0029] Perform principal component analysis on the multi-omics data of the host intestinal flora to obtain the principal component data of the host intestinal flora;

[0030] The diversity of the host intestinal microbial flora is analyzed based on the host intestinal flora Shannon index data and the host intestinal flora principal component data, thereby obtaining the host intestinal microbial flora abundance map;

[0031] The intestinal microbial flora abundance level is divided according to the host intestinal microbial flora abundance map, thereby obtaining the intestinal microbial flora abundance level data;

[0032] Based on the intestinal microbial flora abundance level data, the host intestinal flora multi-omics data is marked with the host intestinal flora abundance level, thereby obtaining the host intestinal multi-omics data.

[0033] The present invention can reveal the relationship between the host and the microorganisms by combining the host physiological coefficient and the multi-omics data of the microbial flora, which helps to understand how intestinal microorganisms affect the physiological state of the host and provide basic data for subsequent research. Shannon index (Shannon index, etc.) analysis can quantify the diversity of the host intestinal flora. This helps to assess the health status of the intestinal microbial community. Higher diversity is usually associated with host health. Principal component analysis helps to reduce the dimension and visualize complex data, making it easier to identify the main influencing factors, which can help researchers understand the similarities and differences between samples and provide intuitive data support for subsequent statistical analysis. Combining the Shannon index and principal component data for diversity analysis can comprehensively evaluate the structure of the microbial community, which can reveal the relative abundance distribution of intestinal microorganisms and help identify the changes of specific microorganisms under different host conditions. According to the microbial abundance map, the intestinal microorganisms can be graded and managed, which helps to understand the health level of the microbial community, clarify potential abnormal conditions, and guide clinical intervention. By marking the host intestinal flora multi-omics data with abundance grade data, the research results can be systematized, providing a basis for subsequent clinical applications or health interventions. This step enhances the availability and practicality of the data.

[0034] Optionally, step S2 is specifically:

[0035] Step S21: extracting the host intestinal microbial flora abundance characteristics based on the host intestinal multi-omics data, thereby obtaining the host intestinal microbial flora abundance data;

[0036] Step S22: performing microbial spatial abundance statistics based on the host intestinal microbial flora abundance data, thereby obtaining intestinal spatial data with high microbial abundance and intestinal spatial data with low microbial abundance;

[0037] Step S23: integrating the intestinal spatial distribution of microbial flora based on the intestinal spatial data of high microbial abundance and the intestinal spatial data of low microbial abundance, thereby obtaining intestinal distribution data of microbial flora;

[0038] Step S24: performing intestinal microbial metabolic pathway analysis based on the intestinal microbial flora distribution data, thereby obtaining intestinal microbial metabolic pathway data;

[0039] Step S25: performing intestinal microbial community multi-omics spatial fusion on the microbial community multi-omics data according to the intestinal microbial metabolic pathway data, thereby obtaining microbial community fusion data.

[0040] The present invention extracts the abundance characteristics of the microbial flora by analyzing the multi-omics data of the host intestine, and can obtain specific abundance data. This step enables researchers to clearly identify and quantify the presence of different microorganisms, helping to understand their role in host physiology and health. Spatial statistics of the abundance data can distinguish high-abundance and low-abundance areas of microorganisms. This analysis can reveal the spatial distribution characteristics of microorganisms in the intestine and provide a basis for understanding the ecological niche and function of the microbial community. Integrating high-abundance and low-abundance intestinal spatial data helps to construct a spatial distribution model of the microbial community, which can help researchers to more comprehensively understand the ecological relationship of microorganisms in the intestine and their interaction with the host. Based on the spatial distribution data of microorganisms, metabolic pathway analysis can reveal the metabolic functions of different microorganisms, which helps to identify the specific roles of microorganisms in host metabolism, nutrient absorption and disease defense, and promote the development of microbiome science towards functional research. Spatial fusion of multi-omics data through intestinal microbial metabolic pathway data can integrate information at different levels to form a comprehensive microbial community map, which helps to enhance the comprehensiveness of the data.

[0041] Optionally, step S3 specifically includes:

[0042] Step S31: extracting host physiological characteristics from the host intestinal multi-omics data to obtain host intestinal physiological data;

[0043] Step S32: integrating the intestinal metabolite concentration characteristics according to the host intestinal physiological data, thereby obtaining intestinal metabolite concentration data;

[0044] Step S33: performing frequency domain conversion on the intestinal metabolite concentration data to obtain an intestinal metabolite concentration spectrum;

[0045] Step S34: performing frequency component classification based on the intestinal metabolite concentration spectrum to obtain metabolite concentration primary frequency component data and metabolite concentration secondary frequency component data;

[0046] Step S35: dividing the intestinal metabolite concentration spectrum into the host normal metabolic physiological cycle according to the metabolite concentration main frequency component data, thereby obtaining the host normal metabolic physiological cycle data; dividing the intestinal metabolite concentration spectrum into the host abnormal metabolic physiological cycle according to the metabolite concentration secondary frequency component data, thereby obtaining the host abnormal metabolic physiological cycle data;

[0047] Step S36: performing host intestinal physiological pattern recognition on the host intestinal physiological data based on the host normal metabolic physiological cycle data, thereby obtaining the host intestinal normal physiological pattern data; performing host intestinal physiological pattern recognition on the host intestinal physiological data based on the host abnormal metabolic physiological cycle data, thereby obtaining the host intestinal abnormal physiological pattern data;

[0048] Step S37: performing physiological pattern time series merging on the host intestinal normal physiological pattern data and the host intestinal abnormal physiological pattern data, thereby obtaining the host intestinal physiological pattern data.

[0049] By integrating different omics data (such as the genome, transcriptome, and metabolome), the present invention can provide a comprehensive understanding of the host's physiological state, providing a solid foundation for subsequent analysis. This step can reveal the interaction between the host and intestinal microorganisms and help understand health and disease states. Integrating different physiological characteristics can more accurately depict the concentration changes of metabolites in the intestine, which provides an important basis for assessing intestinal health and disease risk and can help identify key metabolites. Frequency domain analysis helps identify the changing patterns of metabolites at different time points, allowing researchers to better understand the dynamic characteristics of metabolic activity and reveal potential physiological or pathological mechanisms. By classifying frequency components, it is possible to distinguish between normal and abnormal metabolic patterns, helping to identify key metabolic components, which is of great value in studying specific metabolic pathways or pathological states. By dividing the physiological cycle, a deeper understanding of the host's metabolic changes can be achieved, and the metabolic patterns under specific physiological states can be identified, providing a theoretical basis for personalized medicine and intervention. Pattern recognition can reveal the behavioral characteristics of the intestine under different physiological states and help to construct a physiological model of the host intestine. This is of great significance for early diagnosis and therapeutic intervention of diseases. Through time series merging, data from different physiological patterns can be integrated to form a more comprehensive physiological characteristic map. This will help improve the overall understanding of the host intestinal physiological state and provide data support for subsequent clinical applications.

[0050] Optionally, step S4 is specifically:

[0051] Step S41: extracting microbial-physiological correlation features based on the host intestinal multi-omics data to obtain microbial-physiological correlation data;

[0052] Step S42: performing high-dimensional omics semantic embedding on the microbial flora fusion data according to the microbial-physiological correlation data, thereby obtaining microbial flora dimensionality reduction data;

[0053] Step S43: extracting physiological pattern features from the host intestinal physiological pattern data, thereby obtaining the host intestinal physiological pattern feature data;

[0054] Step S44: performing joint feature matrix integration on the host intestinal physiological pattern feature data and the microbial flora dimensionality reduction data to obtain a physiological-microbial flora joint feature matrix;

[0055] Step S45: performing Spearman correlation analysis based on the physiological-microbial community joint feature matrix to obtain physiological-microbial community correlation data;

[0056] Step S46: Construct a multivariate regression model based on the physiological-microbial correlation data to obtain a regression model of intestinal physiological pattern influencing factors, and calculate the physiological pattern influencing factors based on the regression model of intestinal physiological pattern influencing factors to obtain the host intestinal physiological pattern influencing factors.

[0057] By extracting the correlations between microorganisms and host physiological characteristics, the present invention can identify the impact of key microbial communities and their metabolites on host physiological status, providing basic data for subsequent analysis, facilitating the discovery of potential biomarkers, and providing a basis for early disease diagnosis or health research. Dimensionality reduction can reduce data complexity, remove redundant information, make subsequent analysis more efficient, improve data visualization, and make the relationship between microorganisms and physiological characteristics more intuitive and easy to understand. Extracting the characteristics of host intestinal physiological patterns can provide a deeper understanding of intestinal health status and its interaction with microbial communities, identifying physiological characteristics that have a significant impact on host health, and providing a foundation for personalized medicine. Combining microbial dimensionality reduction data with physiological characteristics helps to fully understand the complex relationship between the two and generate a more complete feature representation. This matrix provides a rich information foundation for subsequent statistical analysis and model building. Through correlation analysis, the relationship between physiological characteristics and microbial communities can be quantified, revealing potential association patterns, providing clues for subsequent causal inference, and helping to determine which microorganisms have a significant impact on physiological status. Constructing regression models can reveal the combined impact of multiple microorganisms and physiological characteristics on host physiological status, providing a theoretical basis for regulating intestinal health. The calculation of influencing factors can provide a scientific basis for personalized intervention strategies (such as dietary adjustment, probiotic application, etc.).

[0058] Optionally, step S5 is specifically as follows:

[0059] Step S51: performing matching variable feature selection based on the host intestinal physiological pattern influencing factors to obtain matching variable data;

[0060] Step S52: performing standardization processing on the matching variable data to obtain the matching variable data to be modeled;

[0061] Step S53: performing microbial community hierarchical clustering on the microbial community fusion data to obtain microbial community clustering data;

[0062] Step S54: constructing a host intestinal flora matching model based on the matching variable data to be modeled, the microbial flora clustering data, and the host intestinal physiological data;

[0063] Step S55: performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data using a host intestinal flora matching model, thereby obtaining intestinal flora matching data.

[0064] The present invention can accurately identify key variables related to the intestinal microbiome by analyzing the influencing factors of the host's intestinal physiological pattern. This process helps to reduce the complexity of the data and ensure that subsequent analysis focuses on the most representative and influential variables, thereby improving the interpretability and predictive performance of the model. Standardization processing enables effective comparison of matching variable data of different dimensions and ranges, eliminating the interference of dimensions on the model. This processing not only improves the comparability of the data, but also enhances the consistency of the model when processing different samples, making the final modeling results more reliable. Through hierarchical clustering, microbial flora can be grouped according to their similarities. This process helps researchers identify the relationship between different groups, reveal the structural characteristics of microbial communities, and provide basic data for understanding the interaction between intestinal flora and hosts. After integrating the matching variable data to be modeled, microbial flora clustering data and host intestinal physiological data, the constructed model can comprehensively reflect the characteristics of the host's intestinal flora. This model provides a scientific basis for subsequent intestinal health assessment and personalized intervention, and can help doctors and nutritionists develop more targeted health plans. By analyzing the microbial flora fusion data and host physiological data through the model, the final intestinal flora matching data obtained can reflect the host's intestinal health status. These data not only help with early warning of diseases, but also provide guidance for personalized diet and microbial intervention strategies, and promote the overall health of the host.

[0065] Optionally, the present invention further provides a multi-omics data optimization system for intestinal flora matching, which is used to execute the multi-omics data optimization method for intestinal flora matching as described above, and the multi-omics data optimization system for intestinal flora matching comprises:

[0066] The microbial diversity integration module is used to obtain historical intestinal multi-omics data and classify the intestinal microbial flora based on the historical intestinal multi-omics data to obtain microbial flora multi-omics data; based on the microbial flora multi-omics data, the host intestinal flora diversity is integrated to obtain host intestinal multi-omics data;

[0067] A multi-omics data fusion module is used to perform microbial flora intestinal distribution statistics based on the host intestinal multi-omics data, thereby obtaining microbial flora intestinal distribution data, and to perform multi-omics data fusion on the microbial flora multi-omics data according to the microbial flora intestinal distribution data, thereby obtaining microbial flora fusion data;

[0068] An intestinal physiological pattern classification module is used to extract host intestinal physiological features based on host intestinal multi-omics data to obtain host intestinal physiological data, and to classify host intestinal physiological patterns based on the host intestinal physiological data to obtain host intestinal physiological pattern data;

[0069] The physiological pattern influencing factor calculation module is used to perform high-dimensional omics semantic embedding on the microbial flora fusion data to obtain the microbial flora dimensionality reduction data; the physiological pattern influencing factor is calculated based on the host intestinal physiological pattern data and the microbial flora dimensionality reduction data to obtain the host intestinal physiological pattern influencing factor;

[0070] The intestinal flora matching module is used to match the host intestinal flora with the microbial flora fusion data and the host intestinal physiological data according to the influencing factors of the host intestinal physiological pattern, so as to obtain the intestinal flora matching data.

[0071] The multi-omics data optimization system for intestinal flora matching of the present invention can implement any multi-omics data optimization method for intestinal flora matching of the present invention, and is used to combine the operations and signal transmission media between various modules to complete the multi-omics data optimization method for intestinal flora matching. The internal modules of the system cooperate with each other to improve the accuracy of flora matching prediction.

[0072] Optionally, the present invention further provides a storage medium storing a computer program, wherein when the computer program is executed by a processor, the multi-omics data optimization method for intestinal flora typing as described above is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0074] Figure 1 A schematic flow chart of the steps of the multi-omics data optimization method for intestinal flora typing of the present invention;

[0075] Figure 2 Detailed step flow diagram of step S1 in the present invention;

[0076] Figure 3 Detailed step flow diagram of step S2 in the present invention;

[0077] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0078] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0079] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0080] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0081] To achieve this, please refer to Figures 1 to 3 The present invention provides a multi-omics data optimization method for intestinal flora typing, the method comprising the following steps:

[0082] Step S1: Acquire historical intestinal multi-omics data, and classify intestinal microbial flora based on the historical intestinal multi-omics data, thereby obtaining microbial flora multi-omics data; integrate the diversity of host intestinal flora based on the microbial flora multi-omics data, thereby obtaining host intestinal multi-omics data;

[0083] In this example, historical intestinal microbiome sequencing data were obtained from public databases (such as MG-RAST or EBI). The raw sequencing data were quality controlled and classified using tools such as QIIME or Mothur, and the microorganisms were classified into species, genus, and family levels. Next, the classification results were analyzed for diversity using the vegan package in R software to calculate the Shannon and Simpson diversity indices of the intestinal microbial community. Finally, these diversity data were integrated to form the host's intestinal multi-omics data for subsequent analysis.

[0084] Step S2: performing microbial flora intestinal distribution statistics based on the host intestinal multi-omics data to obtain microbial flora intestinal distribution data, and performing multi-omics data fusion on the microbial flora multi-omics data according to the microbial flora intestinal distribution data to obtain microbial flora fusion data;

[0085] In this embodiment, based on the host intestinal multi-omics data obtained in the previous step, the ggplot2 and pheatmap packages in R are used to generate distribution maps of microorganisms in different intestinal regions (such as the cecum, colon, and rectum). The abundance data of each microbial taxonomic unit in different intestinal parts are counted, and its relative abundance in each region is calculated. Next, a data fusion method (such as principal component analysis PCA) is used to combine the microbial flora multi-omics data with the intestinal distribution data to form microbial flora fusion data for further analysis.

[0086] Step S3: extracting host intestinal physiological characteristics based on the host intestinal multi-omics data to obtain host intestinal physiological data, and classifying the host intestinal physiological pattern based on the host intestinal physiological data to obtain host intestinal physiological pattern data;

[0087] In this embodiment, the host's metabolomics data is analyzed using bioinformatics tools (such as MetaboAnalyst) to extract physiological characteristics related to intestinal function, such as the concentration of short-chain fatty acids, intestinal pH, and the level of immune markers. These physiological characteristics are then clustered to identify different physiological patterns (such as healthy and diseased states). By using the k-means clustering method, the host intestinal physiological pattern data is obtained, laying the foundation for the calculation of the pattern influencing factors in the next step.

[0088] Step S4: performing high-dimensional omics semantic embedding on the microbial flora fusion data to obtain microbial flora dimensionality reduction data; calculating the physiological pattern influencing factors based on the host intestinal physiological pattern data and the microbial flora dimensionality reduction data to obtain the host intestinal physiological pattern influencing factors;

[0089] In this embodiment, a deep learning model (such as Autoencoder or t-SNE) is used to perform high-dimensional semantic embedding on the microbial flora fusion data to achieve data dimensionality reduction. In this way, the main features can be extracted and the dimensions can be reduced, making the data more suitable for subsequent analysis. Subsequently, combined with the host intestinal physiological pattern data, the factors affecting the host physiological pattern are calculated. For example, the microbial community characteristics related to intestinal health are identified through regression analysis to generate host intestinal physiological pattern influencing factor data.

[0090] Step S5: performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data according to the influencing factors of the host intestinal physiological pattern, thereby obtaining intestinal flora matching data.

[0091] In this example, a machine learning algorithm (such as a random forest or support vector machine) is used to classify the fusion microbial flora data and the host physiological data based on factors influencing the host's intestinal physiological patterns, thereby achieving matching of the host's intestinal flora. The accuracy of the model is evaluated by cross-validation, and finally matching data for different intestinal microbial communities are generated.

[0092] Optionally, step S1 specifically includes:

[0093] Step S11: acquiring historical intestinal multi-omics data and performing data cleaning on the historical intestinal multi-omics data to obtain intestinal multi-omics data to be classified;

[0094] In this embodiment, it is first necessary to obtain historical intestinal multi-omics data from public databases (such as MG-RAST or SRA). The acquired data usually contains information such as genome, transcriptome, and proteome. Next, data cleaning is performed, using tools such as FastQC for quality assessment, removing low-quality sequences and contamination sequences, and using Trimmomatic for sequence trimming to obtain high-quality intestinal multi-omics data to be classified. Finally, the cleaned data set is saved for subsequent analysis.

[0095] Step S12: extracting gene sequencing data and host physiological coefficients from the intestinal multi-omics data to be classified, thereby obtaining intestinal gene sequencing data and host physiological coefficient data;

[0096] In this example, cleaned gene sequencing data was extracted and OTUs (operational taxonomic units) were extracted using bioinformatics tools (such as QiiME or USEARCH). Simultaneously, host physiological parameters, such as weight, BMI, and dietary habits, were collected from the intestinal multi-omics data to be classified and organized into structured data tables using Excel or Python scripts. Data format consistency and integrity were ensured for subsequent analysis.

[0097] Step S13: classifying the microbial flora according to the intestinal gene sequencing data, thereby obtaining microbial flora classification data;

[0098] In this example, bioinformatics software (such as QIIME2 or Mothur) is used to classify the microbial flora based on the extracted intestinal gene sequencing data. By classifying the OTUs and comparing them to existing databases (such as Greengenes or Silva), the classification information of the microorganisms is obtained to form a flora classification data, including the abundance information of each microorganism. The diversity of the flora can be visualized by drawing an abundance histogram or a PCA plot.

[0099] Step S14: performing microbial flora multi-omics classification integration on the intestinal multi-omics data to be classified according to the microbial flora classification data, thereby obtaining microbial flora multi-omics data;

[0100] In this example, the "vegan" package in R was used to integrate the microbial flora classification data with the intestinal multi-omics data to be classified. By cross-analyzing the different omics data, an integrated dataset was generated, containing the abundance of each microorganism in different samples and other relevant omics information. This ultimately generated a microbial flora multi-omics dataset, facilitating subsequent analysis and model building.

[0101] Step S15: Integrate the diversity of the host intestinal flora according to the host physiological coefficient data and the microbial flora multi-omics data, thereby obtaining the host intestinal multi-omics data.

[0102] In this example, the influence of host factors on intestinal microbial diversity was assessed using statistical methods (e.g., linear models or multiple regression analysis) by combining host physiological coefficient data with microbial community multi-omics data. The "phyloseq" package in R was used to calculate diversity indices (e.g., Shannon or Simpson indices) and generate diversity maps to analyze the relationship between host physiological characteristics and intestinal microbial community diversity. Ultimately, host intestinal multi-omics data were generated to facilitate further research and application.

[0103] Optionally, step S13 is specifically as follows:

[0104] Step S131: performing 16S rRNA sequence splicing on the intestinal gene sequencing data to obtain gene sequence splicing data;

[0105] In this embodiment, the original 16S rRNA sequence obtained by intestinal gene sequencing is spliced ​​using bioinformatics tools such as PEAR (Paired-End Assembler for 16S rRNA Sequencing) or Flash. The input data should include forward and reverse sequencing read files (such as .fastq format). During the splicing process, splicing parameters such as maximum splicing length and minimum overlap length are set to ensure that the spliced ​​sequence is complete and high quality. After splicing is completed, the generated output is the gene sequence splicing data, which is usually stored in a new .fastq file for subsequent analysis.

[0106] Step S132: removing redundant sequences based on the gene sequence splicing data, thereby obtaining amplicon sequence variant data;

[0107] In this example, redundant sequences are removed from the spliced ​​16S rRNA sequences using tools such as UCLUST or VSEARCH. The spliced ​​sequences can be imported into the software and a similarity threshold (e.g., 97%) is set for redundancy removal. The output is amplicon sequence variants (ASVs), which represent different microbial populations. The abundance information for each variant is retained and a file containing unique sequences and their abundance is generated, providing basic data for subsequent analysis.

[0108] Step S133: obtaining microbial gene sequence reference data, and performing gene sequence similarity calculation on the microbial gene sequence reference data and amplicon sequence variant data, thereby obtaining gene sequence similarity data;

[0109] In this example, microbial gene sequence reference data is downloaded from a public database (such as Greengenes or Silva). Using tools such as BLAST or USEARCH, the amplicon sequence variants are compared with the reference database and the similarity between the sequences is calculated. A similarity threshold (such as 90%) is set to ensure that only high-quality matches are retained. The output data will include similarity information between each amplicon variant and the reference sequence, and the generated gene sequence similarity data file can be used for subsequent species classification.

[0110] Step S134: performing microbial species classification on the amplicon sequence variant data according to the gene sequence similarity data, thereby obtaining microbial species classification data;

[0111] In this example, microbial species classification was performed using the R package "phyloseq" or "dada2" based on the obtained gene sequence similarity data. Specifically, the amplicon sequence variant data was combined with the similarity data and assigned to species using the nearest neighbor method (Neighbor-Joining) or other classification algorithms. The final output was microbial species classification data, including the species name and classification information for each amplicon sequence, forming a species table that can be used for analysis.

[0112] Step S135: performing microbial species relative abundance calculation on the microbial species classification data, thereby obtaining microbial species relative abundance data;

[0113] In this example, the obtained microbial species classification data, combined with the abundance information of each species included, was used to calculate the relative abundance of each microorganism in the sample. Specifically, the calculation formula is: the abundance of each species divided by the total abundance to obtain the relative abundance percentage. This calculation can be performed using Excel or R programming languages, and the output is a table containing the relative abundance of each species, facilitating subsequent analysis and visualization.

[0114] Step S136: dividing the microbial species classification data into microbial flora according to the microbial species relative abundance data, thereby obtaining microbial flora classification data.

[0115] In this example, microbial species classification data is divided into microbial communities based on relative abundance data using the "vegan" package in the R language or other statistical analysis tools. Species abundance patterns are then divided into distinct microbial communities using cluster analysis (e.g., hierarchical clustering or K-means clustering). The output is microbial community classification data, which includes the composition and characteristics of each sample, providing data support for further ecological analysis and health research.

[0116] Optionally, step S15 is specifically as follows:

[0117] Based on the host physiological coefficient data and the microbial flora multi-omics data, the host flora multi-omics association is performed to obtain the host intestinal flora multi-omics data;

[0118] In this embodiment, the host's physiological coefficient data (such as BMI, age, gender, etc.) and microbial community multi-omics data (such as 16S rRNA sequencing data and metabolomics data) are collected. Using multiple linear regression analysis or random forest model, the host physiological coefficient is associated with microbial community characteristics (such as species abundance, α diversity index) to identify which physiological factors may affect the host's intestinal microbial composition. Finally, a comprehensive data set containing the physiological characteristics and related microbial community data of each host sample is output.

[0119] Perform Shannon Index diversity analysis on the multi-omics data of the host intestinal flora to obtain the Shannon Index data of the host intestinal flora;

[0120] In this example, Shannon-Wiener Index diversity analysis was performed on the obtained multi-omics data of the host gut microbiota. First, the abundance of each microbial species in each sample was calculated using the formula: H' = -Σ(p_i * log(p_i)), where p_i is the relative abundance of each species. The vegan package in R can be used for this calculation, and the Shannon-Wiener Index data for each sample is output to assess the diversity level of the gut microbiome.

[0121] Perform principal component analysis on the multi-omics data of the host intestinal flora to obtain the principal component data of the host intestinal flora;

[0122] In this example, principal component analysis (PCA) was performed using the obtained multi-omics data of the host intestinal flora. After the abundance data were standardized, they were converted to principal component space by PCA, and the first few principal components were extracted to explain the variance of the data. The Python sklearn library was used for analysis, and a principal component plot was drawn. The principal component data was output to facilitate understanding of the structure and variation of the microbial community in subsequent analysis.

[0123] The diversity of the host intestinal microbial flora is analyzed based on the host intestinal flora Shannon index data and the host intestinal flora principal component data, thereby obtaining the host intestinal microbial flora abundance map;

[0124] In this example, the diversity of the host intestinal microbial flora was analyzed by combining the Shannon Index data and the principal component data. Cluster analysis methods (such as hierarchical clustering) were used to identify different microbial communities and generate abundance maps to show the differences in microbial abundance between different samples. The ggplot2 package in the R language can be used for visualization and output a microbial community abundance distribution map for each sample.

[0125] The intestinal microbial flora abundance level is divided according to the host intestinal microbial flora abundance map, thereby obtaining the intestinal microbial flora abundance level data;

[0126] In this embodiment, the microbial community of each sample is divided into different levels of abundance based on the generated intestinal microbial flora abundance map. Specific abundance thresholds (such as high, medium, and low) are set and samples are divided into different levels based on the abundance distribution. Customized grading standards can be used, for example, considering an abundance of ≥50% as high abundance, 20%-50% as medium abundance, and <20% as low abundance, to ultimately generate abundance level data.

[0127] Based on the intestinal microbial flora abundance level data, the host intestinal flora multi-omics data is marked with the host intestinal flora abundance level, thereby obtaining the host intestinal multi-omics data.

[0128] In this example, based on the obtained gut microbial abundance level data, this level information is tagged back into the host gut microbiome multi-omics data. A new field is added to the dataset for each sample to identify its abundance level, thereby forming a comprehensive dataset containing host physiological information, microbial community characteristics, and their abundance levels. This data can be used for subsequent statistical analysis and model building to explore the relationship between abundance level and host health status.

[0129] Optionally, step S2 is specifically:

[0130] Step S21: extracting the host intestinal microbial flora abundance characteristics based on the host intestinal multi-omics data, thereby obtaining the host intestinal microbial flora abundance data;

[0131] In this embodiment, DNA is extracted from the host's fecal sample, and the whole genome sequence of the microbial community is obtained by high-throughput sequencing technology (such as Illumina sequencing). Then, the sequencing data is processed using bioinformatics tools (such as QIIME or Mothur) to extract the abundance data of each microbial species. Specifically, the analysis results include the relative abundance percentage of each microorganism in the sample, forming an abundance matrix for subsequent analysis. It is also possible to extract abundance features such as defined abundance levels and relative abundance of intestinal microbial flora from the host intestinal multi-omics data.

[0132] Step S22: performing microbial spatial abundance statistics based on the host intestinal microbial flora abundance data, thereby obtaining intestinal spatial data with high microbial abundance and intestinal spatial data with low microbial abundance;

[0133] In this embodiment, the obtained abundance data are statistically analyzed using a statistical method. The intestine can be divided into several regions (such as the cecum, ileum, and colon), and the abundance of microorganisms in each region is statistically analyzed. By using software (such as the vegan package in R), the diversity of microbial species in each region can be calculated, and the microorganisms can be divided into two groups: high abundance (such as> 10% relative abundance) and low abundance (such as <1% relative abundance), thereby obtaining microbial high abundance intestinal spatial data and low abundance intestinal spatial data. Each intestinal region can also be divided into two groups of high abundance and low abundance according to the defined abundance level.

[0134] Step S23: integrating the intestinal spatial distribution of microbial flora based on the intestinal spatial data of high microbial abundance and the intestinal spatial data of low microbial abundance, thereby obtaining intestinal distribution data of microbial flora;

[0135] In this example, data on high and low microbial abundances are integrated and the spatial distribution of the intestinal microbial community is visualized using a geographic information system (GIS) or related visualization tools (such as Circos plots or heat maps). Specifically, this can include cluster analysis of microbial abundance in different regions to further reveal the spatial distribution patterns of microorganisms in the intestine and generate a spatial distribution map of the intestinal microbial community.

[0136] Step S24: performing intestinal microbial metabolic pathway analysis based on the intestinal microbial flora distribution data, thereby obtaining intestinal microbial metabolic pathway data;

[0137] In this example, metabolic pathway analysis was performed based on the obtained microbial community distribution data. The KEGG database or MetaCyc database can be used, combined with abundance data, to identify metabolic pathways active in specific intestinal regions. For example, PICRUSt software can be used to predict microbial functions and, combined with abundance information, generate a graphical representation of the main metabolic pathways in each region. This will help understand how microorganisms affect host metabolic function.

[0138] Step S25: performing intestinal microbial community multi-omics spatial fusion on the microbial community multi-omics data according to the intestinal microbial metabolic pathway data, thereby obtaining microbial community fusion data.

[0139] In this embodiment, the microbial abundance data is spatially fused with other omics data (such as transcriptome, metabolome) according to the intestinal microbial metabolic pathway data. Multi-omics integration tools (such as MOFA or mixOmics) can be used to map different omics data (microorganisms, transcriptomes, metabolomes) to the same spatial framework based on the intestinal spatial characteristics of the intestinal microbial metabolic pathway data and the multi-omics data of the microbial flora to generate comprehensive microbial flora fusion data. This data will help understand the complex interactions of the intestinal microbial community in the host's health or disease state.

[0140] Optionally, step S3 specifically includes:

[0141] Step S31: extracting host physiological characteristics from the host intestinal multi-omics data to obtain host intestinal physiological data;

[0142] In this embodiment, a feature selection algorithm (such as random forest or LASSO regression) is used to identify important features related to the host's intestinal health status. For example, by analyzing transcriptome data, the expression levels of genes related to the intestinal flora are extracted to obtain the host's intestinal physiological data. The accuracy and stability of the extracted features are ensured by cross-validation.

[0143] Step S32: integrating the intestinal metabolite concentration characteristics according to the host intestinal physiological data, thereby obtaining intestinal metabolite concentration data;

[0144] In this example, the obtained host physiological data is matched with a known metabolite database, and the data is integrated using machine learning methods (such as support vector machines) to obtain concentration characteristics of intestinal metabolites. For example, the concentrations of short-chain fatty acids (SCFAs) such as acetate and propionate are integrated, and their correlation with host physiological characteristics is evaluated using regression analysis methods to obtain integrated metabolite concentration data.

[0145] Step S33: performing frequency domain conversion on the intestinal metabolite concentration data to obtain an intestinal metabolite concentration spectrum;

[0146] In this example, a fast Fourier transform (FFT) algorithm was used to perform frequency domain conversion on intestinal metabolite concentration data. This process converts the time domain signal into a frequency domain signal, thereby extracting the intestinal metabolite concentration spectrum. For example, FFT processing was performed on the acetic acid concentration data to obtain its spectral characteristics, revealing the patterns of metabolite changes over time.

[0147] Step S34: performing frequency component classification based on the intestinal metabolite concentration spectrum to obtain metabolite concentration primary frequency component data and metabolite concentration secondary frequency component data;

[0148] In this example, waveform analysis techniques were applied to the obtained intestinal metabolite concentration spectrum to decompose the spectrum data into primary and secondary frequency components. Wavelet transform analysis can be used to identify data with different frequency components. For example, wavelet transform can identify the most significant frequency components in the spectrum, determining that the primary frequency component is 0.1 Hz and the secondary frequency component is 0.05 Hz.

[0149] Step S35: dividing the intestinal metabolite concentration spectrum into the host normal metabolic physiological cycle according to the metabolite concentration main frequency component data, thereby obtaining the host normal metabolic physiological cycle data; dividing the intestinal metabolite concentration spectrum into the host abnormal metabolic physiological cycle according to the metabolite concentration secondary frequency component data, thereby obtaining the host abnormal metabolic physiological cycle data;

[0150] In this embodiment, based on the main frequency component, a time series analysis method (such as an ARIMA model) is used to divide the intestinal metabolite concentration spectrum into periods to identify the host's normal metabolic physiological period. For the secondary frequency component, the host's abnormal metabolic physiological period is identified by an anomaly detection algorithm (such as Isolation Forest). This process can help determine the time frame of normal and abnormal metabolic states. The main frequency component data is used to divide the host's normal metabolic physiological period, and it can also be classified according to the 24-hour biological rhythm. For abnormal metabolism, the secondary frequency component is used to perform event-driven period division. The sliding window technology is used to determine the switching point between normal and abnormal conditions by analyzing the peaks and valleys in the time series. This process will produce two important data sets: host normal metabolic physiological period data and abnormal metabolic physiological period data.

[0151] Step S36: performing host intestinal physiological pattern recognition on the host intestinal physiological data based on the host normal metabolic physiological cycle data, thereby obtaining the host intestinal normal physiological pattern data; performing host intestinal physiological pattern recognition on the host intestinal physiological data based on the host abnormal metabolic physiological cycle data, thereby obtaining the host intestinal abnormal physiological pattern data;

[0152] In this embodiment, cluster analysis (such as K-means or DBSCAN) is applied to perform pattern recognition on the host's intestinal physiological data. The host's intestinal normal physiological pattern is identified based on the normal metabolic physiological cycle data, and the abnormal physiological pattern is identified by anomaly detection. For example, a machine learning algorithm (such as a support vector machine, a decision tree) is used to perform pattern recognition on intestinal physiological data. Normal physiological patterns can be characterized by cluster analysis; abnormal physiological patterns need to be identified in combination with physiological events (such as dietary changes, disease states). A training set and a test set can be constructed, and the reliability of the model can be ensured by cross-validation, and the host's intestinal normal and abnormal physiological pattern data can be obtained.

[0153] Step S37: performing physiological pattern time series merging on the host intestinal normal physiological pattern data and the host intestinal abnormal physiological pattern data, thereby obtaining the host intestinal physiological pattern data.

[0154] In this embodiment, the identified normal and abnormal physiological patterns are temporally merged. Dynamic time warping (DTW) technology is used to align and merge physiological patterns within different time periods to generate host intestinal physiological pattern data. This process comprehensively considers the host's physiological changes and provides a holistic physiological pattern view, facilitating subsequent analysis and application.

[0155] Optionally, step S4 is specifically:

[0156] Step S41: extracting microbial-physiological correlation features based on the host intestinal multi-omics data to obtain microbial-physiological correlation data;

[0157] In this example, bioinformatics analysis was performed using machine learning algorithms (e.g., random forests and support vector machines) to integrate different omics data and extract correlation features between microorganisms and host physiological status. Specific microbial species (e.g., bifidobacteria or lactic acid bacteria) were selected and the correlation between their abundance and host physiological indicators (e.g., intestinal pH, short-chain fatty acid concentration, etc.) was calculated, ultimately forming a dataset containing microbial characteristics and their correlation with host physiological status.

[0158] Step S42: performing high-dimensional omics semantic embedding on the microbial flora fusion data according to the microbial-physiological correlation data, thereby obtaining microbial flora dimensionality reduction data;

[0159] In this embodiment, microbial features related to physiological patterns are selected for the extracted microbial-physiological association data, and high-dimensional embedding technology (such as Word2Vec or BERT) is applied to reduce the dimensionality of the microbial flora fusion data. The feature vectors of each microorganism are quantized, and similar features are clustered together through cluster analysis (such as K-means), reducing the dimension while retaining important information. The final output of the microbial flora dimensionality reduction data will provide a clear feature representation for subsequent analysis, which will help to better understand the functional characteristics of microorganisms. In this process, the data is first standardized to eliminate the dimensional effect, and then a deep learning algorithm is applied to extract potential features, thereby compressing the high-dimensional features into a low-dimensional representation. This dimensionality-reduced data not only reduces the computational complexity, but also retains key information about the structure of the microbial community.

[0160] Step S43: extracting physiological pattern features from the host intestinal physiological pattern data, thereby obtaining the host intestinal physiological pattern feature data;

[0161] In this example, feature extraction is performed on host intestinal physiological pattern data (such as inflammatory markers and metabolic levels). Time series analysis can be used to analyze changes in physiological data at different time points, combined with machine learning algorithms (such as support vector machines and decision trees) for feature selection. This will generate a set of physiological pattern feature data that reveals changes in the host's physiological state and its potential relationship with the microbiome.

[0162] Step S44: performing joint feature matrix integration on the host intestinal physiological pattern feature data and the microbial flora dimensionality reduction data to obtain a physiological-microbial flora joint feature matrix;

[0163] In this embodiment, the host intestinal physiological pattern feature data and the microbial flora dimensionality reduction data are integrated to form a physiological-flora joint feature matrix. The microbial flora dimensionality reduction data are integrated with the host intestinal physiological pattern feature data to form a physiological-flora joint feature matrix. Data fusion techniques (such as weighted average or principal component fusion) can be used to process data from different sources. For example, different weights are given to each feature to ensure that important features dominate the joint matrix, and ultimately a feature matrix that comprehensively reflects the correlation between microorganisms and host physiological characteristics is formed.

[0164] Step S45: performing Spearman correlation analysis based on the physiological-microbial community joint feature matrix to obtain physiological-microbial community correlation data;

[0165] In this example, a Spearman correlation analysis was performed using a combined physiological-microbial feature matrix. By calculating the rank correlation between different features, the degree of association between physiological and microbial characteristics can be identified. The Spearman method is particularly well-suited for nonparametric data and can reveal potential nonlinear relationships. The results generate physiological-microbial correlation data, providing statistical support for subsequent model construction.

[0166] Step S46: Construct a multivariate regression model based on the physiological-microbial correlation data to obtain a regression model of intestinal physiological pattern influencing factors, and calculate the physiological pattern influencing factors based on the regression model of intestinal physiological pattern influencing factors to obtain the host intestinal physiological pattern influencing factors.

[0167] In this embodiment, a multivariate regression model is constructed based on physiological-microbial correlation data to identify factors affecting intestinal physiological patterns. The model can use methods such as linear regression or ridge regression to consider the multiple effects of physiological characteristics and microbial community characteristics. After the regression model is trained, the accuracy and stability of the model are evaluated by cross-validation. Finally, based on the coefficients of the regression model, the influencing factors of the host intestinal physiological pattern can be calculated.

[0168] Optionally, step S5 is specifically as follows:

[0169] Step S51: performing matching variable feature selection based on the host intestinal physiological pattern influencing factors to obtain matching variable data;

[0170] In this example, key factors influencing the host's intestinal physiological patterns, such as age, gender, dietary habits, and medication use, are identified. Next, statistical analysis methods (e.g., random forest or LASSO regression) are used to perform feature selection on these factors to identify the variables with the greatest predictive power. Assuming that the analysis reveals that "dietary fiber intake" and "intestinal movement frequency" are key features, these two variables will be taken into consideration when constructing the matching variable dataset.

[0171] Step S52: performing standardization processing on the matching variable data to obtain the matching variable data to be modeled;

[0172] In this example, the obtained matching variable data is standardized to eliminate dimensional differences between different variables. The Z-score standardization method can be used, which calculates the mean and standard deviation of each variable and then performs a transformation. Specifically, for a variable value x, the standardized result is z = \frac{x - \mu}{\sigma}, where \mu is the mean and \sigma is the standard deviation. After standardization, the values ​​of the matching variable data to be modeled are evenly distributed, which facilitates subsequent modeling.

[0173] Step S53: performing microbial community hierarchical clustering on the microbial community fusion data to obtain microbial community clustering data;

[0174] In this embodiment, for microbial flora data, the similarity between samples is first calculated, usually using the Bray-Curtis distance or Jaccard index. Then, a hierarchical clustering algorithm (such as UPGMA) is used to construct a dendrogram to cluster similar microbial populations together. By selecting an appropriate clustering threshold (e.g., a distance less than 0.5), the samples are divided into multiple classes to facilitate subsequent analysis. The resulting cluster data can reflect the similarity of the microbial composition in different samples.

[0175] Step S54: constructing a host intestinal flora matching model based on the matching variable data to be modeled, the microbial flora clustering data, and the host intestinal physiological data;

[0176] In this example, the standardized matching variable data, microbial flora clustering data, and host intestinal physiological data are combined to select an appropriate machine learning algorithm (such as random forest or support vector machine) to construct a matching model. The training data must include labeled samples to improve model accuracy. The goal of the final model is to predict the intestinal flora status of new samples and its relationship with host physiological characteristics.

[0177] Step S55: performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data using a host intestinal flora matching model, thereby obtaining intestinal flora matching data.

[0178] In this example, the constructed model is used to predict new microbial flora fusion data and host physiological data to obtain host intestinal flora matching data. The microbial data and physiological characteristics of a new sample are input, and the model generates a matching result. For example, the model may output a sample as "balanced" or "imbalanced," and based on the results, corresponding dietary and lifestyle adjustment strategies can be studied.

[0179] Optionally, the present invention further provides a multi-omics data optimization system for intestinal flora matching, which is used to execute the multi-omics data optimization method for intestinal flora matching as described above, and the multi-omics data optimization system for intestinal flora matching comprises:

[0180] The microbial diversity integration module is used to obtain historical intestinal multi-omics data and classify the intestinal microbial flora based on the historical intestinal multi-omics data to obtain microbial flora multi-omics data; based on the microbial flora multi-omics data, the host intestinal flora diversity is integrated to obtain host intestinal multi-omics data;

[0181] A multi-omics data fusion module is used to perform microbial flora intestinal distribution statistics based on the host intestinal multi-omics data, thereby obtaining microbial flora intestinal distribution data, and to perform multi-omics data fusion on the microbial flora multi-omics data according to the microbial flora intestinal distribution data, thereby obtaining microbial flora fusion data;

[0182] An intestinal physiological pattern classification module is used to extract host intestinal physiological features based on host intestinal multi-omics data to obtain host intestinal physiological data, and to classify host intestinal physiological patterns based on the host intestinal physiological data to obtain host intestinal physiological pattern data;

[0183] The physiological pattern influencing factor calculation module is used to perform high-dimensional omics semantic embedding on the microbial flora fusion data to obtain the microbial flora dimensionality reduction data; the physiological pattern influencing factor is calculated based on the host intestinal physiological pattern data and the microbial flora dimensionality reduction data to obtain the host intestinal physiological pattern influencing factor;

[0184] The intestinal flora matching module is used to match the host intestinal flora with the microbial flora fusion data and the host intestinal physiological data according to the influencing factors of the host intestinal physiological pattern, so as to obtain the intestinal flora matching data.

[0185] The multi-omics data optimization system for intestinal flora matching of the present invention can implement any multi-omics data optimization method for intestinal flora matching of the present invention, and is used to combine the operations and signal transmission media between various modules to complete the multi-omics data optimization method for intestinal flora matching. The internal modules of the system cooperate with each other to improve the accuracy of flora matching prediction.

[0186] Optionally, the present invention further provides a storage medium storing a computer program, wherein when the computer program is executed by a processor, the multi-omics data optimization method for intestinal flora typing as described above is implemented.

[0187] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0188] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A multi-omics data optimization method for intestinal flora typing, characterized in that: The following steps are involved: Step S1: obtaining historical intestinal multi-omics data, and classifying intestinal microbial flora according to the historical intestinal multi-omics data, thereby obtaining microbial flora multi-omics data; integrating the diversity of host intestinal flora according to the microbial flora multi-omics data, thereby obtaining host intestinal multi-omics data; Step S2: performing microbial flora intestinal distribution statistics based on the host intestinal multi-omics data to obtain microbial flora intestinal distribution data, and performing multi-omics data fusion on the microbial flora multi-omics data according to the microbial flora intestinal distribution data to obtain microbial flora fusion data; Step S3: extracting host intestinal physiological characteristics according to the host intestinal multi-omics data to obtain host intestinal physiological data, and classifying host intestinal physiological patterns based on the host intestinal physiological data to obtain host intestinal physiological pattern data; Step S4: high-dimensional omics semantic embedding is performed on the microbial flora fusion data to obtain the microbial flora dimensionality reduction data; physiological pattern influencing factors are calculated based on the host intestinal physiological pattern data and the microbial flora dimensionality reduction data to obtain the host intestinal physiological pattern influencing factors; Step S5: performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data according to the influencing factors of the host intestinal physiological pattern, so as to obtain intestinal flora matching data.

2. The multi-omics data optimization method for intestinal flora typing according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: acquiring historical intestinal multi-omics data, and performing data cleaning on the historical intestinal multi-omics data, thereby obtaining intestinal multi-omics data to be classified; Step S12: extracting gene sequencing data and host physiological coefficients from the intestinal multi-omics data to be classified, thereby obtaining intestinal gene sequencing data and host physiological coefficient data; Step S13: classifying the microbial flora according to the intestinal gene sequencing data, thereby obtaining microbial flora classification data; Step S14: performing microbial flora multi-omics classification integration on the intestinal multi-omics data to be classified according to the microbial flora classification data, thereby obtaining microbial flora multi-omics data; Step S15: Integrate the diversity of the host intestinal flora according to the host physiological coefficient data and the microbial flora multi-omics data, so as to obtain the host intestinal multi-omics data.

3. The multi-omics data optimization method for intestinal flora typing according to claim 2, characterized in that: Step S13 is specifically as follows: Step S131: performing 16S rRNA sequence splicing on the intestinal gene sequencing data to obtain gene sequence splicing data; Step S132: removing redundant sequences according to the gene sequence splicing data, thereby obtaining amplicon sequence variant data; Step S133: obtaining microbial gene sequence reference data, and performing gene sequence similarity calculation on the microbial gene sequence reference data and amplicon sequence variant data, thereby obtaining gene sequence similarity data; Step S134: classifying the amplicon sequence variant data into microbial species according to the gene sequence similarity data, thereby obtaining microbial species classification data; Step S135: performing microbial species relative abundance calculation on the microbial species classification data, thereby obtaining microbial species relative abundance data; Step S136: dividing the microbial species classification data into microbial flora according to the microbial species relative abundance data, thereby obtaining microbial flora classification data.

4. The multi-omics data optimization method for intestinal flora typing according to claim 2, characterized in that: Step S15 is specifically as follows: According to the host physiological coefficient data and the multi-omics data of the microbial flora, the multi-omics association of the host flora is performed to obtain the multi-omics data of the host intestinal flora; The Shannon index diversity analysis of the multi-omics data of the host intestinal flora was performed to obtain the Shannon index data of the host intestinal flora; Perform principal component analysis on the multi-omics data of the host intestinal flora to obtain the principal component data of the host intestinal flora; The diversity of the host's intestinal microbial flora is analyzed according to the host's intestinal flora fragrance index data and the host's intestinal flora principal component data, so as to obtain the host's intestinal microbial flora abundance map; According to the intestinal microbial flora abundance map of the host, the intestinal microbial flora abundance level is divided into different levels, thereby obtaining the intestinal microbial flora abundance level data; Based on the intestinal microbial flora abundance level data, the host intestinal flora multi-omics data are labeled with the host intestinal flora abundance level to obtain the host intestinal multi-omics data.

5. The multi-omics data optimization method for intestinal flora typing according to claim 1, characterized in that: Step S2 is specifically as follows: Step S21: extracting the abundance characteristics of the host intestinal microbial flora according to the host intestinal multi-omics data, thereby obtaining the abundance data of the host intestinal microbial flora; Step S22: performing microbial spatial abundance statistics according to the intestinal microbial flora abundance data of the host, thereby obtaining intestinal spatial data with high microbial abundance and intestinal spatial data with low microbial abundance; Step S23: integrating the intestinal spatial distribution of microbial flora according to the intestinal spatial data of high microbial abundance and the intestinal spatial data of low microbial abundance, thereby obtaining intestinal distribution data of microbial flora; Step S24: analyzing the intestinal microbial metabolic pathway based on the intestinal distribution data of the microbial flora, thereby obtaining intestinal microbial metabolic pathway data; Step S25: performing intestinal microbial flora multi-omics spatial fusion on the microbial flora multi-omics data according to the intestinal microbial metabolic pathway data, thereby obtaining microbial flora fusion data.

6. The multi-omics data optimization method for intestinal flora typing according to claim 1, characterized in that: Step S3 is specifically as follows: Step S31: extracting host physiological characteristics from the host intestinal multi-omics data to obtain host intestinal physiological data; Step S32: integrating the intestinal metabolite concentration characteristics according to the host intestinal physiological data, thereby obtaining intestinal metabolite concentration data; Step S33: performing frequency domain conversion on the intestinal metabolite concentration data to obtain the intestinal metabolite concentration spectrum; Step S34: classifying the frequency components according to the intestinal metabolite concentration spectrum, thereby obtaining the metabolite concentration main frequency component data and the metabolite concentration secondary frequency component data; Step S35: dividing the intestinal metabolite concentration spectrum into the host normal metabolic physiological cycle according to the metabolite concentration main frequency component data, thereby obtaining the host normal metabolic physiological cycle data; According to the sub-frequency component data of metabolite concentration, the intestinal metabolite concentration spectrum is divided into the host abnormal metabolic physiological cycle, so as to obtain the host abnormal metabolic physiological cycle data; Step S36: performing host intestinal physiological pattern recognition on the host intestinal physiological data based on the host normal metabolic physiological cycle data, thereby obtaining the host intestinal normal physiological pattern data; Performing host intestinal physiological pattern recognition on host intestinal physiological data according to the host abnormal metabolic physiological cycle data, thereby obtaining the host intestinal abnormal physiological pattern data; Step S37: performing physiological pattern time series merging on the host intestinal normal physiological pattern data and the host intestinal abnormal physiological pattern data, so as to obtain the host intestinal physiological pattern data.

7. The multi-omics data optimization method for intestinal flora typing according to claim 1, characterized in that: Step S4 is specifically as follows: Step S41: extracting microbial-physiological correlation features based on the host intestinal multi-omics data, thereby obtaining microbial-physiological correlation data; Step S42: performing high-dimensional omics semantic embedding on the microbial flora fusion data according to the microbial-physiological association data, thereby obtaining microbial flora dimensionality reduction data; Step S43: extracting physiological pattern features from the host intestinal physiological pattern data, thereby obtaining the host intestinal physiological pattern feature data; Step S44: integrating the host intestinal physiological pattern feature data and the microbial flora dimensionality reduction data into a joint feature matrix, thereby obtaining a physiological-flora joint feature matrix; Step S45: performing Spearman correlation analysis based on the physiological-microbial community joint feature matrix to obtain physiological-microbial community correlation data; Step S46: construct a multivariate regression model based on the physiological-microbial correlation data to obtain a regression model of intestinal physiological pattern influencing factors, and calculate the physiological pattern influencing factors based on the regression model of intestinal physiological pattern influencing factors to obtain the host intestinal physiological pattern influencing factors.

8. The multi-omics data optimization method for intestinal flora typing according to claim 1, characterized in that: Step S5 is specifically as follows: Step S51: performing matching variable feature selection according to the host intestinal physiological pattern influencing factors to obtain matching variable data; Step S52: standardizing the matching variable data to obtain the matching variable data to be modeled; Step S53: performing microbial flora hierarchical clustering on the microbial flora fusion data, thereby obtaining microbial flora clustering data; Step S54: constructing a host intestinal flora matching model according to the matching variable data to be modeled, the microbial flora clustering data and the host intestinal physiological data; Step S55: performing host intestinal flora matching on the microbial flora fusion data and the host intestinal physiological data through the host intestinal flora matching model, thereby obtaining intestinal flora matching data.

9. A multi-omics data optimization system for intestinal flora matching, characterized in that: Used to execute the multi-omics data optimization method for intestinal flora matching according to claim 1, the multi-omics data optimization system for intestinal flora matching comprises: The microbiota diversity integration module is used to obtain historical intestinal multi-omics data, and classify the intestinal microbiota based on the historical intestinal multi-omics data, so as to obtain the microbiota multi-omics data; integrate the diversity of the host intestinal microbiota based on the microbiota multi-omics data, so as to obtain the host intestinal multi-omics data; A multi-omics data fusion module is used to perform microbial flora intestinal distribution statistics based on the host intestinal multi-omics data, so as to obtain microbial flora intestinal distribution data, and to perform multi-omics data fusion on the microbial flora multi-omics data according to the microbial flora intestinal distribution data, so as to obtain microbial flora fusion data; An intestinal physiological pattern classification module is used to extract host intestinal physiological characteristics based on the host intestinal multi-omics data to obtain host intestinal physiological data, and to classify the host intestinal physiological patterns based on the host intestinal physiological data to obtain host intestinal physiological pattern data; The physiological pattern influencing factor calculation module is used to perform high-dimensional omics semantic embedding on the microbial flora fusion data, so as to obtain the microbial flora dimensionality reduction data; the physiological pattern influencing factor is calculated based on the host intestinal physiological pattern data and the microbial flora dimensionality reduction data, so as to obtain the host intestinal physiological pattern influencing factor; The intestinal flora matching module is used to match the host intestinal flora with the microbial flora fusion data and the host intestinal physiological data according to the influencing factors of the host intestinal physiological pattern, so as to obtain the intestinal flora matching data.

10. A storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the multi-omics data optimization method for intestinal flora matching according to any one of claims 1 to 8 is implemented.

Images