An artificial intelligence-based microecological regulation effect analysis method and system

Abstract

The present application provides a kind of microecological regulation effect analysis method and system based on artificial intelligence, it is related to artificial intelligence and biological information analysis technical field, the present application is by collecting regulation process basic data, historical regulation sample data, to be analyzed sample flora data and host response data, based on historical regulation sample establishment artificial intelligence microecological response model, obtain the reference state of flora recovery, risk flora inhibition, inflammation improvement and mucosal barrier improvement;Combining the time sequence variation of to-be-analyzed sample, analyze flora synergistic recovery, microecological improvement back swing and flora and host response consistency, obtain microecological regulation effect evaluation result, and accordingly generate the candidate input dose adjustment strategy of next regulation cycle, realize the iterative update of regulation effect, improve the dynamics and integrity of microecological regulation effect analysis.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and bioinformatics analysis technology, specifically to a method and system for analyzing the effects of microecological regulation based on artificial intelligence. Background Technology

[0002] Gut microbiota regulation typically involves altering the gut microbiome composition of research subjects through measures such as probiotics, prebiotics, dietary active ingredients, or traditional Chinese medicine extracts. The regulatory effects are then analyzed using relative abundance of the gut microbiota, inflammatory markers, and mucosal barrier indicators. Current analyses of gut microbiota regulation effectiveness often rely on gut microbiota testing results before and after regulation, or host response testing results, neglecting to consider the continuous changes between different sampling times.

[0003] However, gut microbiota regulation exhibits significant temporal variations, and the recovery of associated microbiota and the suppression of risk-associated microbiota may have different response rates. Furthermore, some study subjects may experience a decrease in the degree of microbiota improvement after reaching a certain level of improvement. Simultaneously, there may be differences in response time and trends between changes in microbiota status and inflammatory or mucosal barrier responses. Based solely on a single test result or a single type of indicator, it is difficult to distinguish between different regulatory states such as insufficient synergistic recovery of the microbiota, post-improvement relapse, and inconsistencies between microbiota and host responses. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for analyzing the effects of microecological regulation based on artificial intelligence, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for analyzing the effects of micro-ecological regulation based on artificial intelligence, comprising the following steps:

[0007] Step 1: Monitor the intestinal microecological regulation of the research subjects using probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts, collect basic data of the regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed, and determine the improvement-related microbial community and risk-related microbial community based on the historical regulation sample data;

[0008] Step 2: Based on historical regulation sample data, establish an artificial intelligence model training sample set, use a long short-term memory neural network for training, and establish an artificial intelligence microecological response model; obtain the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators, and determine the reference state intervals for improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement.

[0009] Step 3: Based on the microbial community data of the sample to be analyzed, the improved associated microbial community, the risk associated microbial community, the effective recovery reference relative abundance of the improved associated microbial community, and the effective inhibition reference relative abundance of the risk associated microbial community, obtain the recovery amount of the improved associated microbial community and the inhibition amount of the risk associated microbial community at each sampling time; based on the recovery amount of the improved associated microbial community, the inhibition amount of the risk associated microbial community, the reference state interval of the improved associated microbial community, the reference state interval of the risk associated microbial community, and the basic data of the regulation process at each sampling time, obtain the microbial community co-recovery coefficient;

[0010] Step 4: Based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, obtain the comprehensive microecological improvement amount at each sampling time, and conduct a time-series change analysis on the comprehensive microecological improvement amount at each sampling time to obtain the microecological swing coefficient.

[0011] Step 5: Based on the host response data, effective improvement reference values ​​of inflammatory indicators, and effective improvement reference values ​​of mucosal barrier indicators of the samples to be analyzed, obtain the amount of inflammation improvement and mucosal barrier improvement at each sampling time, and obtain the amount of host improvement at each sampling time based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time; based on the comprehensive microecological improvement, host improvement, comprehensive microecological improvement reference state interval, host improvement reference state interval, and basic data of the regulatory process at each sampling time, obtain the consistency coefficient between the microbiota and the host response.

[0012] Step 6: Based on the synergistic recovery coefficient of the microbial community, the microecological swing coefficient, and the consistency coefficient between the microbial community and the host response, a comprehensive comparative analysis is conducted to obtain the evaluation results of the microecological regulation effect; and candidate dosage adjustment strategies for the next regulation cycle are generated, and the regulation effect of the next regulation cycle is iteratively updated.

[0013] An artificial intelligence-based microecological regulation effect analysis system includes:

[0014] The microecological data acquisition module is used to monitor research subjects that use probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts to regulate the intestinal microecology. It collects basic data of the regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed. Based on the historical regulation sample data, it identifies the microbial communities associated with improvement and the microbial communities associated with risk.

[0015] The microecological response modeling module is used to establish an artificial intelligence model training sample set based on historical regulation sample data. It uses a long short-term memory neural network for training to establish an artificial intelligence microecological response model. It also obtains the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators. Furthermore, it determines the reference state intervals for improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement.

[0016] The microbial community synergistic recovery analysis module is used to obtain the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time based on the microbial community data of the sample to be analyzed, improved associated microbial communities, risk associated microbial communities, effective recovery reference relative abundance of improved associated microbial communities, and effective inhibition reference relative abundance of risk associated microbial communities. Based on the recovery amount of improved associated microbial communities, the inhibition amount of risk associated microbial communities, the reference state interval of improved associated microbial communities, the reference state interval of risk associated microbial communities, and the basic data of the regulation process at each sampling time, the microbial community synergistic recovery coefficient is obtained.

[0017] The microecological swing analysis module is used to obtain the comprehensive microecological improvement at each sampling time based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, and to perform time-series change analysis on the comprehensive microecological improvement at each sampling time to obtain the microecological swing coefficient.

[0018] The microbiota-host response consistency analysis module is used to obtain the amount of inflammation improvement and mucosal barrier improvement at each sampling time based on the host response data, effective improvement reference values ​​of inflammatory indicators, and effective improvement reference values ​​of mucosal barrier indicators of the sample to be analyzed. It also obtains the amount of host improvement at each sampling time based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time. Based on the comprehensive microecological improvement, host improvement, comprehensive microecological improvement reference state interval, host improvement reference state interval, and basic data of the regulatory process at each sampling time, the microbiota-host response consistency coefficient is obtained.

[0019] The regulation effect evaluation and update module is used to conduct a comprehensive comparative analysis based on the microbial community synergistic recovery coefficient, the microecological swing coefficient, and the microbial community-host response consistency coefficient to obtain the microecological regulation effect evaluation results; and generate candidate dosage adjustment strategies for the next regulation cycle, and perform iterative updates of the regulation effect for the next regulation cycle.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] This invention continuously acquires microbial community data, inflammatory response data, and mucosal barrier response data at multiple sampling times before and after microecological regulation, and establishes an artificial intelligence microecological response model by combining historical regulation samples. This transforms the analysis of microecological regulation effects from a single point-in-time judgment to a continuous time-series analysis, which can simultaneously reflect the changes in the recovery of improved associated microbial communities, the suppression of risk-associated microbial communities, and the host response during the regulation process. This improves the completeness of the characterization of the dynamic process of microecological regulation and reduces the bias in state judgment caused by relying solely on the last detection result.

[0022] This invention also further distinguishes different states such as insufficient regulation, post-improvement relapse, inconsistency between microbiota and host response, and continuous effective regulation by jointly analyzing the synchronous recovery of improved associated microbiota and risk associated microbiota, the changes after the comprehensive microecological improvement state reaches its maximum improvement, and the consistency between microbiota changes and host inflammation and mucosal barrier response. This makes the evaluation of microecological regulation effect not only reflect whether improvement has occurred, but also whether the improvement is stable and whether microbiota changes match the host response, thus improving the refinement and interpretability of the evaluation results of microecological regulation effect.

[0023] This invention also differentiates the subsequent administration dose based on the evaluation results of the microecological regulation effect, and continues to analyze the microecological regulation effect after obtaining the microbial community data and host response data for the next regulation cycle, so that the current regulation effect can form a continuous feedback relationship with the subsequent regulation process; different follow-up treatment methods are adopted for different states such as continuous effectiveness, post-improvement reversal, insufficient regulation, and inconsistency between microbial community and host response, thereby realizing iterative updating of candidate administration doses and improving the matching degree between the adjustment of microecological regulation parameters and the current actual response state. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0025] Figure 2 This is a schematic diagram of the overall system flow of the present invention. Detailed Implementation

[0026] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0027] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0028] This invention combines basic data on the microecological regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed. It uses an artificial intelligence microecological response model to establish a reference for historical effective regulation states, and sequentially analyzes the recovery of improved related microbial communities, the suppression of risk-related microbial communities, the state changes after comprehensive microecological improvement, and the matching between microbial community response and host response. This makes the effect of microecological regulation no longer limited to a single sampling time or a single detection index, but can form a continuous analysis from multiple aspects such as the degree of microbial community recovery, the maintenance of the improved state, and the consistency of host response. Furthermore, it can form a candidate dosage adjustment strategy for the next regulation cycle, improving the completeness of the microecological regulation effect analysis and the pertinence of subsequent regulation.

[0029] Please see Figure 1 This invention provides a method for analyzing the effects of microecological regulation based on artificial intelligence, the specific steps of which include:

[0030] Step 1: Monitor the intestinal microecological regulation of research subjects using probiotics, prebiotics, dietary active ingredients, or traditional Chinese medicine extracts. Collect basic data on the regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed. Based on the historical regulation sample data, identify the microbial communities associated with improvement and those at risk. Specifically:

[0031] S11. Monitor the regulation process of intestinal microecology in research subjects using probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts, and collect basic data of the regulation process, including: type of microecological regulation measure, current dose, start time of regulation, sampling time after regulation and total duration of regulation analysis time window;

[0032] Specifically, the following steps are taken: Record the type of probiotics, prebiotics, dietary active ingredients, or traditional Chinese medicine extracts actually used by the current research subjects to determine the type of microecological regulation measures; obtain the start time of regulation by reading the administration records; obtain the current administration dose D0 using the existing dose measurement method corresponding to the current regulation method; for solid regulation materials, weigh the actual administration mass using an electronic balance; for liquid regulation materials, read the actual administration volume using a pipette, graduated cylinder, or applicator; for probiotic preparations measured according to the number of live bacteria, obtain the corresponding administration dose based on the live bacteria count indicated on the preparation label and the actual administration volume.

[0033] The sampling times after regulation are determined according to the pre-set fixed sampling interval. The start time of regulation is taken as the start time of the regulation analysis time window, and the last sampling time is taken as the end time of the regulation analysis time window. The total duration T of the regulation analysis time window is obtained based on the time difference between the two.

[0034] S12. Retrieve data from historical microecological regulation processes that are the same type as the current microecological regulation measures, collect historical regulation sample data, including: historical microbial community time-series data, historical inflammatory response time-series data, historical mucosal barrier response time-series data, and historical actual regulation effect results, and determine the improvement-related microbial communities and risk-related microbial communities based on the historical microbial community time-series data and historical actual regulation effect results.

[0035] Specifically, the process involves: retrieving historical control samples of the same type as the current microecological control measures; obtaining the relative abundance of each bacterial community category at multiple sampling times before and after microecological control for each historical control sample, as historical microbial community time-series data; obtaining the detection values ​​of inflammatory indicators at the corresponding sampling times, as historical inflammatory response time-series data; obtaining the detection values ​​of mucosal barrier indicators at the corresponding sampling times, as historical mucosal barrier response time-series data; and obtaining the actual control effect results that have been validated for each historical control sample.

[0036] Based on the relationship between the relative abundance changes of each microbial community category after microecological regulation and the historical actual regulation results, the microbial community categories that show a stable positive relationship with the improvement of the actual regulation effect are identified as improvement-related microbial communities, and the microbial community categories that show a stable negative relationship with the improvement of the actual regulation effect are identified as risk-related microbial communities.

[0037] S13. Periodically monitor the gut microbiota of the research subjects and collect microbiota data of the samples to be analyzed, including: improving the relative abundance of related microbiota before regulation. The relative abundance of associated bacterial communities improved at each sampling time point after regulation. Relative abundance of risk-associated microbial communities before regulation Relative abundance of risk-associated bacterial communities at each sampling time after regulation ;

[0038] Specifically, gut microbiota samples were collected from the study subjects at various sampling times before and after the implementation of the current microecological regulation measures. Valid microbial sequences were obtained using 16S rRNA amplicon sequencing, and species annotation methods were used to classify and identify these sequences. Based on the improved associated microbiota and the risk-associated microbiota, the number of valid sequences for each microbiota was counted, and the proportion of valid sequences for each microbiota to the total number of valid microbial sequences in the current sample was determined as the relative abundance of that microbiota. The relative abundance of the improved associated microbiota before regulation was then obtained. The relative abundance of associated bacterial communities improved at each sampling time point after regulation. Relative abundance of risk-associated microbial communities before regulation Relative abundance of risk-associated bacterial communities at each sampling time after regulation ;in, This represents the relative abundance of the i-th improved associated bacterial community before microecological regulation. This represents the relative abundance of the i-th improved associated bacterial community at the t-th sampling time after regulation. This represents the relative abundance of the j-th risk-associated bacterial community before microecological regulation. This represents the relative abundance of the j-th risk-associated bacterial community at the t-th sampling time after regulation.

[0039] S14. Periodically monitor the host response status of the research subjects and collect host response data of the samples to be analyzed, including: pre-modulation inflammatory marker detection values. Inflammatory marker values ​​at various sampling times after regulation Pre-regulation mucosal barrier index detection values Mucosal barrier index values ​​at various sampling times after regulation ;

[0040] Specifically, blood samples were collected from the study subjects at various sampling times before and after microecological regulation. The collected blood samples underwent coagulation and centrifugation to separate serum. Enzyme-linked immunosorbent assay (ELISA) was used to detect inflammation-related cytokines in the serum, and the pre-regulation inflammatory marker values ​​were obtained. Inflammatory marker values ​​at each sampling time after regulation ;

[0041] Colonic tissue was collected from parallel study samples at various sampling times before and after microecological regulation. Proteins were extracted from the colonic tissue, and the expression levels of tight junction proteins were detected using Western blot. The results were then quantified to obtain the pre-regulation mucosal barrier index values. Mucosal barrier index values ​​at various sampling times after regulation ;

[0042] in, This represents the detection value of the k-th inflammatory marker before microecological regulation. Inflammatory markers include: tumor necrosis factor α, interleukin 1β, interleukin 6, and interleukin 10. This represents the detection value of the k-th inflammatory marker at the t-th sampling time after regulation; This represents the detection value of the h-th mucosal barrier index before microecological regulation; This represents the detection value of the h-th mucosal barrier index at the t-th sampling time after regulation.

[0043] This invention records the current dosage, start time of regulation, sampling times after regulation, and regulation analysis time window for different microecological regulation measures. It also retrieves historical microbial communities, inflammatory responses, mucosal barrier responses, and actual regulatory effects data corresponding to the same type of microecological regulation measure. Combined with 16S rRNA gene amplicon sequencing to obtain the relative abundance of microbial communities, and simultaneously obtains host response data through the detection of inflammation-related cytokines and tight junction proteins, it enables the current regulation process, historical effective regulation processes, microbial community changes, and host changes to establish a correspondence according to a unified regulation process. This provides a data foundation with clear temporal relationships and biological meaning for subsequent artificial intelligence modeling and dynamic effect analysis.

[0044] Step 2: Based on historical regulation sample data, establish an artificial intelligence model training sample set, using a long short-term memory neural network for training, and establish an artificial intelligence microecological response model; obtain the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk-associated microbiota, the effective improvement reference detection values ​​of inflammatory indicators, and the effective improvement reference detection values ​​of mucosal barrier indicators, and determine the reference state intervals for improved associated microbiota, risk-associated microbiota, comprehensive microecological improvement, and host improvement. Specifically:

[0045] S21. Based on historical regulation sample data, the historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data corresponding to each historical regulation sample are arranged sequentially according to the sampling time before and after microecological regulation, according to the sample number. The historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data are standardized using the Z-score standardization method, and the mean and standard deviation of each type of data after Z-score standardization are recorded as Z-score standardization parameters. The standardized historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data are used as time-series input data to establish an artificial intelligence model training sample set.

[0046] S22. Using a stratified random sampling method, the training sample set of the artificial intelligence model is divided into a training set and a validation set. Then, a long short-term memory neural network is used for training. The relative abundance of improved associated bacteria, relative abundance of risk associated bacteria, inflammatory marker detection values, and mucosal barrier index detection values ​​corresponding to consecutive sampling times are used as time-series inputs. The relative abundance of improved associated bacteria, relative abundance of risk associated bacteria, inflammatory marker detection values, and mucosal barrier index detection values ​​corresponding to the next sampling time are used as training targets. Backpropagation and gradient descent methods are used to update the network parameters. The network parameters are determined based on the prediction error of the validation set to establish an artificial intelligence microecological response model.

[0047] S23. Using a label matching screening method, historical control samples corresponding to effective control categories are selected based on historical actual control effect results. Z-score standardization parameters are used to standardize the relative abundance of improved associated bacteria, relative abundance of risk associated bacteria, inflammatory marker detection values, and mucosal barrier marker detection values ​​of each effective historical control sample at each sampling time before and after microecological control. The standardized data is then input into the artificial intelligence microecological response model to obtain the predicted sequences corresponding to each effective historical control sample. Z-score inverse standardization is used to restore each predicted sequence to its corresponding original data scale according to the Z-score standardization parameters. The prediction results at the last sampling time of each predicted sequence are extracted, and the effective recovery reference relative abundance of the i-th improved associated bacteria is obtained using the average value calculation method. The effective suppression of the j-th risk-associated bacterial community as a reference relative abundance Reference value for effective improvement of the kth inflammatory marker The h-th mucosal barrier index effectively improved the reference detection value. .

[0048] S24. Based on the predicted sequences restored to the original data scale corresponding to each effective historical regulation sample, as well as the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators, the degree of recovery of improved associated microbiota, the degree of inhibition of risk associated microbiota, the degree of improvement of comprehensive microecology, and the degree of improvement of host at each sampling time are normalized. The quartile statistical method is used to determine the reference state intervals of improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement, respectively.

[0049] This invention organizes historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data according to sample number and sampling time. After Z-score standardization, it uses a long short-term memory neural network to learn the state change relationship between consecutive sampling times. This enables the model to utilize multiple types of biological data that are time-dependent in the process of microecological regulation. Furthermore, by regulating historical samples of effective categories, it obtains reference states for microbial community recovery, microbial community inhibition, inflammation improvement, and mucosal barrier improvement, and restores the model prediction results to the original data scale. Thus, it provides a historical and effective reference basis corresponding to the actual detection data scale for calculating the degree of improvement of the current research object in subsequent aspects.

[0050] Step 3: Based on the microbial community data of the sample to be analyzed, the improved associated microbial communities, the risk associated microbial communities, the effective recovery reference relative abundance of the improved associated microbial communities, and the effective inhibition reference relative abundance of the risk associated microbial communities, obtain the recovery amount of the improved associated microbial communities and the inhibition amount of the risk associated microbial communities at each sampling time. Based on the recovery amount of the improved associated microbial communities, the inhibition amount of the risk associated microbial communities, the reference state interval of the improved associated microbial communities, the reference state interval of the risk associated microbial communities, and the basic data of the regulation process at each sampling time, obtain the microbial community co-recovery coefficient. Specifically:

[0051] S31, Based on improving the relative abundance of associated bacterial communities before regulation The relative abundance of associated bacterial communities improved at each sampling time point after regulation. And the effective recovery reference relative abundance for improving associated microbial communities Using a normalized boundary constraint method, the community recovery ratio of the i-th improved associated bacterial community at the t-th sampling time is calculated. The formula is as follows:

[0052]

[0053] In the formula, Let represent the recovery ratio of the i-th improved associated microbial community at the t-th sampling time, and ε represent a dimensionless minimum positive number to prevent the denominator from being zero.

[0054] S32, Based on the i-th improved associated bacterial community at the t-th sampling time, the bacterial community recovery ratio is... The arithmetic mean method was used to calculate the recovery amount of the improved associated microbial community at the t-th sampling time. The formula is as follows:

[0055]

[0056] In the formula, m represents the improvement in the number of associated bacteria. This represents the amount of improved associated microbial community recovery at the t-th sampling time. The larger the value, the closer the improvement in the overall associated microbial community is to an effective recovery to the reference state. The smaller the value, the lower the overall recovery level of the associated microbial community.

[0057] S33, Based on the relative abundance of risk-associated microbial communities before regulation Relative abundance of risk-associated bacterial communities at each sampling time after regulation Effective suppression of risk-associated bacterial communities as a reference relative abundance Using a normalized boundary constraint method, the community inhibition ratio of the j-th risk-associated bacterial community at the t-th sampling time is calculated. The formula is as follows:

[0058] .

[0059] S34. Based on the community suppression ratio of the j-th risk-associated bacterial community at the t-th sampling time. The arithmetic mean method was used to calculate the inhibition level of the risk-associated bacterial community at the t-th sampling time. The formula is as follows:

[0060]

[0061] In the formula, n represents the number of risk-associated microbial communities. This represents the amount of risk-associated bacterial community suppression at the t-th sampling time. The larger the value, the closer the overall risk-associated microbiota is to the effective inhibition reference state. The smaller the value, the lower the overall level of inhibition of the risk-associated microbiota.

[0062] This invention compares the relative abundance changes of improved associated bacterial communities at each sampling time with their effective recovery reference state, and compares the relative abundance changes of risk associated bacterial communities at each sampling time with their effective inhibition reference state. Then, it summarizes the calculation results for multiple improved associated bacterial communities and multiple risk associated bacterial communities. This allows changes in different bacterial community categories and at different relative abundance scales to be converted into recovery and inhibition levels with unified significance, avoiding the judgment of the overall microecological effect based solely on the increase or decrease of the abundance of a single bacterial community. This improves the ability to characterize the overall recovery state of improved associated bacterial communities and the overall inhibition state of risk associated bacterial communities.

[0063] Step three of the present invention also includes:

[0064] S35, Improved recovery of associated microbial communities based on the t-th sampling time. Inhibition of risk-associated microbiota The time-series end-value extraction method was used to extract the recovery amount of improved associated bacteria and the inhibition amount of risk associated bacteria at the last sampling time, and the last recovery amount of improved associated bacteria (UFm) and the last inhibition amount of risk associated bacteria (UHm) were obtained respectively. Based on the recovery amount of improved associated bacteria, the inhibition amount of risk associated bacteria, and the reference state intervals of improved associated bacteria and risk associated bacteria at each sampling time, the interval matching and sequential retrieval method was used to perform interval matching according to the sampling time order, and the time tF when improved associated bacteria first entered the corresponding reference state interval and the time tH when risk associated bacteria first entered the corresponding reference state interval were obtained respectively.

[0065] S36. Based on the last recovery amount UFm of the improved associated microbial community, the last inhibition amount UHm of the risk associated microbial community, the time tF when the improved associated microbial community first enters the corresponding reference state interval, the time tH when the risk associated microbial community first enters the corresponding reference state interval, and the total duration T of the regulation analysis time window, the microbial community co-recovery coefficient XTF is calculated, as follows:

[0066]

[0067] In the formula, XTF represents the synergistic recovery coefficient of the microbial community; the larger the XTF, the more coordinated the degree of recovery of the improved associated microbial community and the degree of inhibition of the risk associated microbial community, and the smaller the time difference between the two entering the corresponding reference state interval; the smaller the XTF, the lower the coordination between the degree of recovery and inhibition or the larger the time difference between the two entering the corresponding reference state interval.

[0068] This invention extracts the recovery and inhibition states of both improved and risk-associated bacterial communities at the last sampling time, and determines the time when each type of community first enters the corresponding reference state interval. By combining the actual recovery degree at the end of the regulation phase with the time relationship of reaching the effective state during the regulation process, the evaluation of community synergistic recovery can not only reflect whether the two types of communities ultimately reach the corresponding improved state, but also reflect the synchronicity of the recovery of improved bacterial communities and the inhibition of risk-associated bacterial communities in terms of response time. This reduces the problem of failing to identify asynchronous responses between the two types of communities when only the last community detection results are used.

[0069] Step 4: Based on the recovery amount of improved associated microbial communities and the inhibition amount of risk-associated microbial communities at each sampling time, obtain the comprehensive microecological improvement amount at each sampling time, and perform time-series variation analysis on the comprehensive microecological improvement amount at each sampling time to obtain the microecological swing coefficient. Specifically:

[0070] S41, Improved recovery of associated microbial communities based on the t-th sampling time. Inhibition of risk-associated microbiota The arithmetic mean method was used to calculate the comprehensive microecological improvement at the t-th sampling time. The formula is as follows:

[0071]

[0072] In the formula, This represents the overall improvement in the microecological environment at the t-th sampling time. The larger the value, the higher the overall improvement in the microecology at the current sampling time. The smaller the value, the lower the degree of improvement in the overall microecology.

[0073] S42. Comprehensive microecological improvement based on the t-th sampling time. The method employs extreme value retrieval and time-series difference statistics to compare the magnitudes of various comprehensive microecological improvements and performs difference analysis on adjacent sampling times after the maximum improvement time to obtain the maximum comprehensive microecological improvement. The corresponding maximum improvement time and the comprehensive microecological improvement at the last sampling time. The number of sampling intervals where the overall microecological improvement decreased after the point of maximum improvement. And the total number of sampling intervals after the time of maximum improvement. .

[0074] S43, Based on the maximum comprehensive microecological improvement Comprehensive microecological improvement at the last sampling time The number of sampling intervals where the overall microecological improvement decreased after the point of maximum improvement. Total number of sampling intervals after the time of maximum improvement The formula for calculating the microecological swing coefficient SHB is as follows:

[0075]

[0076] In the formula, SHB represents the microecological swing coefficient; the larger the SHB, the greater the decrease in the overall microecological improvement after reaching the maximum improvement state and the more sampling intervals the decrease occurs, indicating a higher degree of microecological swing; the smaller the SHB, the higher the degree of maintenance of the overall microecological state after reaching the maximum improvement state and a lower degree of microecological swing; when the maximum improvement time corresponding to the maximum overall microecological improvement is the last sampling time, the microecological swing coefficient SHB is determined to be 0.

[0077] This invention combines the recovery state of improved microbial communities and the inhibition state of risk-associated microbial communities at each sampling time to form a continuous comprehensive microecological improvement state. It performs extreme value retrieval and time-series difference analysis on the continuous sampling results after reaching the maximum improvement state. It also considers the degree of improvement decline from the maximum improvement state to the last sampling time and the sampling interval where the decline occurs. This allows the regulation process that has shown significant improvement but then gradually weakens to be distinguished from the regulation process that maintains the improvement state. This improves the ability to identify the phenomenon of "post-improvement reversal" in the microecology and avoids overestimating the actual maintenance effect of the entire regulation cycle based solely on the results of the optimal sampling time.

[0078] Step 5: Based on the host response data, effective improvement reference values ​​for inflammatory markers, and effective improvement reference values ​​for mucosal barrier markers of the samples to be analyzed, obtain the amount of inflammation improvement and mucosal barrier improvement at each sampling time. Then, based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time, obtain the amount of host improvement at each sampling time. Based on the comprehensive microecological improvement, host improvement, comprehensive microecological improvement reference state interval, host improvement reference state interval, and basic data of the regulatory process at each sampling time, obtain the consistency coefficient between the microbiota and the host response. Specifically:

[0079] S51, Based on pre-regulation inflammatory marker detection values Inflammatory marker values ​​at various sampling times after regulation Effective improvement of reference test values ​​for inflammatory markers Using reference distance normalization and boundary constraint methods, the improvement ratio of the k-th inflammatory marker at the t-th sampling time is calculated. The formula is as follows:

[0080]

[0081] In the formula, This represents a very small positive number with the same dimensions as the k-th inflammatory marker detection value, used to prevent the denominator from being zero.

[0082] S52, the improvement ratio of the inflammatory marker based on the k-th inflammatory marker at the t-th sampling time. The arithmetic mean method was used to calculate the amount of inflammation improvement at the t-th sampling time. The formula is as follows:

[0083]

[0084] In the formula, p represents the number of inflammatory markers. This represents the amount of inflammation improvement at the t-th sampling time. The higher the value, the closer the overall inflammation marker is to the reference state for effective improvement. The smaller the value, the lower the degree of improvement in the inflammatory response.

[0085] S53, Based on the detection values ​​of mucosal barrier indicators before regulation Mucosal barrier index values ​​at various sampling times after adjustment Effective improvement of mucosal barrier indicators and reference test values Using reference distance normalization and boundary constraint methods, the improvement ratio of the h-th mucosal barrier index at the t-th sampling time was calculated. The formula is as follows:

[0086]

[0087] In the formula, This represents a very small positive number with the same dimensions as the h-th mucosal barrier index detection value, used to prevent the denominator from being zero.

[0088] S54. Improvement ratio of mucosal barrier index at sampling time t based on the h-th mucosal barrier index. The arithmetic mean method was used to calculate the improvement in the mucosal barrier at the t-th sampling time. The formula is as follows:

[0089]

[0090] In the formula, q represents the number of mucosal barrier indicators. This represents the improvement in the mucosal barrier at the t-th sampling time. The higher the value, the closer the overall mucosal barrier index is to the reference state of effective improvement. The smaller the value, the lower the degree of improvement in the mucosal barrier response.

[0091] This invention uses the effective improvement reference values ​​of inflammatory indicators and mucosal barrier indicators as the basis for host response analysis. It normalizes the inflammatory and mucosal barrier indicators at each sampling time before and after regulation by reference distance, and summarizes them to form the inflammatory improvement state and the mucosal barrier improvement state respectively. This allows different types and detection scales of host response indicators to be uniformly represented according to the degree of effective improvement of the reference state by their distance. This simultaneously reflects the changes in host state in both inflammatory and mucosal barrier responses, reducing the one-sidedness caused by relying solely on a single inflammatory indicator or a single barrier indicator to evaluate the effect of microecological regulation.

[0092] S55, Inflammation improvement based on the t-th sampling time. and the amount of mucosal barrier improvement The arithmetic mean method is used to calculate the host improvement at the t-th sampling time. The formula is as follows:

[0093]

[0094] In the formula, This represents the host improvement at the t-th sampling time. The larger the value, the greater the overall improvement in the inflammatory response and mucosal barrier response. The smaller the value, the lower the degree of improvement in the host response.

[0095] S56. Host improvement amount based on the t-th sampling time. The host improvement value at the last sampling time was extracted using the time-series last value extraction method. The comprehensive improvement in the microecology based on the t-th sampling time. and host improvement amount Pearson correlation analysis was used to perform time-series correlation analysis to obtain the correlation coefficient between the overall improvement in the microecology and the improvement in the host organism. The comprehensive improvement in the microecology based on the t-th sampling time. Host improvement In addition to the reference state intervals for comprehensive microecological improvement and host improvement, interval matching and sequential retrieval methods were used to perform interval matching according to the sampling time sequence, and the timestM and tG for the first time of comprehensive microecological improvement and host improvement were obtained respectively.

[0096] S57, Comprehensive Microecological Improvement Based on the Last Sampling Time Host improvement at the last sampling time The time tM when the comprehensive microecological improvement first enters the corresponding reference state interval; the time tG when the host improvement first enters the corresponding reference state interval; and the correlation coefficient between the comprehensive microecological improvement and the host improvement. Given the total duration T of the regulation analysis time window, the consistency coefficient JSX between the microbial community and the host response is calculated using the following formula:

[0097]

[0098] In the formula, JSX represents the consistency coefficient between the microbial community and the host response; the larger the JSX, the closer the overall microecological improvement is to the host improvement, the smaller the time difference between the two entering the corresponding reference state interval, and the more consistent the change trend within the entire regulatory time window; the smaller the JSX, the lower the consistency between the microbial community response and the host response in one or more of the improvement degree, response time, or change trend.

[0099] This invention integrates the improvement status of inflammation and the improvement status of the mucosal barrier to form the host improvement status at each sampling time, and performs synchronous time-series analysis on the host improvement status and the overall microecological improvement status. It also considers the last degree of improvement, the time relationship of the first entry into their respective reference state intervals, and the change trend within the entire regulation time window. This allows the correspondence between microbial changes and host physiological responses to be comprehensively characterized from three aspects: improvement magnitude, response time, and change trend. This enables the identification of situations where the microbial community has improved but the host response has not improved synchronously, thereby improving the ability of microecological regulation effect analysis to identify differences between microbial community and host response.

[0100] Step Six: Based on the microbial community synergistic recovery coefficient, microecological swing coefficient, and microbial community-host response consistency coefficient, a comprehensive comparative analysis is performed to obtain the evaluation results of the microecological regulation effect; and candidate dosage adjustment strategies for the next regulation cycle are generated, and the regulation effect is iteratively updated for the next regulation cycle. Specifically:

[0101] S61. By setting a predefined threshold for microbial community synergistic recovery (Xth), a microbial ecosystem reversal threshold (Sth), and a threshold for microbial community-host response consistency (Jth), and by comprehensively comparing and analyzing the microbial community synergistic recovery coefficient (XTF), the microbial ecosystem reversal coefficient (SHB), and the microbial community-host response consistency coefficient (JSX) with the microbial community synergistic recovery threshold (Xth), the microbial ecosystem reversal threshold (Sth), and the microbial community-host response consistency threshold (Jth), the evaluation results of the microbial ecosystem regulation effect are obtained, including:

[0102] When the synergistic recovery coefficient XTF < the synergistic recovery threshold Xth, it indicates that the synergistic recovery results of the degree of improvement of the associated microbial community recovery and the degree of inhibition of the risk associated microbial community in the sample to be analyzed have not met the standard, and the sample to be analyzed is judged to be in a state of insufficient regulation.

[0103] When the microbial community synergistic recovery coefficient XTF ≥ microbial community synergistic recovery threshold Xth, and the microecological swing coefficient SHB > microecological swing threshold Sth, it indicates that the synergistic recovery results of the degree of recovery of the improved associated microbial community and the degree of inhibition of the risk associated microbial community in the sample to be analyzed meet the standard, but the overall microecological improvement decreases after reaching the maximum improvement state, and the microecological swing result is unqualified. The sample to be analyzed is judged to be in the swing state after improvement.

[0104] When the microbial community co-recovery coefficient XTF ≥ microbial community co-recovery threshold Xth, the microbial ecosystem swing coefficient SHB ≤ microbial ecosystem swing threshold Sth, and the microbial community-host response consistency coefficient JSX < microbial community-host response consistency threshold Jth, it indicates that the microbial community co-recovery result of the sample to be analyzed meets the standard and the microbial ecosystem swing result is qualified. However, the consistency between the overall microbial ecosystem improvement degree and the host improvement degree in terms of improvement magnitude, response time, or change trend does not meet the standard, and it is determined that the sample to be analyzed is in a state of inconsistent microbial community-host response.

[0105] When the microbial community co-recovery coefficient XTF ≥ microbial community co-recovery threshold Xth, the microbial ecosystem swing coefficient SHB ≤ microbial ecosystem swing threshold Sth, and the microbial community-host response consistency coefficient JSX ≥ microbial community-host response consistency threshold Jth, it indicates that the microbial community co-recovery result of the sample to be analyzed meets the standard, the microbial ecosystem swing result is qualified, and the consistency between the microbial community response and the host response meets the standard, and it is determined that the sample to be analyzed is in a state of continuous and effective regulation.

[0106] S62. Based on the evaluation results of the microecological regulation effect and the current administered dose D0, generate candidate administration dose adjustment strategies for the next regulation cycle, including:

[0107] When the evaluation result of the microecological regulation effect is a state of continuous effective regulation, the current administration dose D0 is maintained, and the current administration dose is used as the candidate administration dose for the next regulation cycle;

[0108] When the evaluation result of the microecological regulation effect is that the state of improvement has returned to normal, the dose D0 is increased by 5% to 10% based on the current dose D0, and the adjusted dose is used as the candidate dose for the next regulation cycle.

[0109] When the evaluation result of the microecological regulation effect is that the regulation is insufficient, the dosage D0 is increased by 10% to 15% based on the current dosage, and the adjusted dosage is used as the candidate dosage for the next regulation cycle.

[0110] When the evaluation result of the microecological regulation effect is that the microbial community and the host response are inconsistent, the current administration dose D0 is maintained, the current microecological regulation process is extended to the next sampling time, and the consistency coefficient between the microbial community and the host response is recalculated after acquiring the microbial community data and the host response data of the newly added sample to be analyzed. The candidate administration dose for the next regulation cycle is determined based on the updated evaluation result of the microecological regulation effect.

[0111] S63. Implement subsequent microecological regulation based on the candidate dose for the next regulation cycle, and after obtaining the microbial community data and host response data of the sample to be analyzed corresponding to the next regulation cycle, recalculate and iteratively update the evaluation results of the microecological regulation effect and the candidate dose for the next regulation cycle.

[0112] Method for obtaining the microbial community co-recovery threshold Xth: The calibration target of the microbial community co-recovery threshold Xth is to determine the critical value of the microbial community co-recovery coefficient that can distinguish between "the state of microbial community co-recovery results meeting the standard" and "the state of microbial community co-recovery results not meeting the standard".

[0113] This invention constructs a historical database of effective microecological regulation, with a sample size of no less than 500 groups, covering different types of microecological regulation measures, different dosages, different regulation analysis time windows, different pre-regulation microbial community states, and different post-regulation microbial community response processes. Each sample group includes historical microbial community time-series data, historical inflammatory response time-series data, historical mucosal barrier response time-series data, and validated historical actual regulation results. Historical regulation samples whose historical actual regulation results belong to the effective regulation category are selected from this database. This sample organization method corresponds to the historical regulation sample data and effective regulation category selection method of this invention.

[0114] Based on the relative abundance of improved associated microbial communities and risk associated microbial communities of each historical effective regulation sample, and according to the calculation methods of microbial community recovery ratio, improved associated microbial community recovery amount, microbial community inhibition ratio, and risk associated microbial community inhibition amount of the present invention, the last recovery amount of improved associated microbial communities, the last inhibition amount of risk associated microbial communities, and the time when the two types of microbial communities first enter the corresponding reference state interval of each historical effective regulation sample are obtained respectively. According to the calculation method of microbial community co-recovery coefficient XTF of the present invention, the historical microbial community co-recovery coefficient corresponding to each historical effective regulation sample is obtained respectively, and all historical microbial community co-recovery coefficients are arranged in ascending order. The historical microbial community co-recovery coefficient corresponding to the 25th position in the sorting result is taken as the lower quartile co-recovery value, and the lower quartile co-recovery value is determined as the microbial community co-recovery threshold Xth.

[0115] The microbial community co-recovery coefficient (XTF) can simultaneously reflect the degree of recovery of the associated microbial community at the last time, the degree of inhibition of the risk-associated microbial community at the last time, and the time difference between the two types of microbial communities reaching the reference state. The value of the microbial community co-recovery coefficient is in the range of 0 to 1. The larger the value, the more sufficient the recovery and inhibition of the two types of microbial communities and the closer the response time. The technical meaning of XTF is that "the larger the value, the more coordinated the degree of recovery and inhibition, and the smaller the time difference to enter the reference state interval." Preferably, the lower quartile co-recovery value of the historical microbial community co-recovery coefficient in the historical effective microecological regulation database is set to 0.55. Therefore, the microbial community co-recovery threshold Xth is preferably 0.55.

[0116] When the microbial community synergistic recovery coefficient (XTF) is less than 0.55, it indicates that the current degree of improvement in the recovery of associated microbial communities, the degree of inhibition of risk-associated microbial communities, or the coordination of their response times has not yet reached the lower quartile level of historical effective regulatory samples, and the microbial community synergistic recovery result is judged to be unsatisfactory. When the microbial community synergistic recovery coefficient (XTF) reaches or exceeds 0.55, it indicates that the current comprehensive synergistic state of microbial community recovery and inhibition has reached the lower quartile level of historical effective regulatory samples, and the microbial community synergistic recovery result is judged to be satisfactory.

[0117] The calibration target of the microecological swing threshold Sth is to determine the critical value of the microecological swing coefficient that can distinguish between "qualified state of microecological swing result" and "unqualified state of microecological swing result".

[0118] By accessing the historical effective microecological regulation database, the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities corresponding to each sampling time are obtained from historical regulation samples whose actual historical regulation effects belong to the category of effective regulation. The comprehensive microecological improvement time series data corresponding to each historical effective regulation sample are obtained according to the calculation method of the comprehensive microecological improvement amount of this invention. Furthermore, the maximum comprehensive microecological improvement amount, the comprehensive microecological improvement amount at the last sampling time, the number of sampling intervals in which the comprehensive microecological improvement amount decreased after the maximum improvement time, and the total number of sampling intervals after the maximum improvement time are obtained.

[0119] According to the calculation method of the microecological swing coefficient SHB, the historical microecological swing coefficients corresponding to each historical effective regulation sample are obtained respectively. All historical microecological swing coefficients are arranged in ascending order. The historical microecological swing coefficients located at the 75th position in the sorting results are taken as the upper quartile swing value, and the upper quartile swing value is determined as the microecological swing threshold Sth.

[0120] Based on the characteristics of the microecological swing coefficient SHB, which is determined by the relative decrease from the maximum improvement state to the last sampling time and the proportion of sampling intervals where a decrease occurs after the maximum improvement time, the microecological swing coefficient is within the range of 0 to 1. The larger the value, the greater the degree of decrease and the more sustained the decrease process after reaching the maximum improvement state; the smaller the value, the higher the degree of maintenance after the maximum improvement state. When the maximum improvement time is located at the last sampling time, the microecological swing coefficient is determined to be 0. Preferably, the upper quartile swing value of the historical microecological swing coefficient in the historical effective microecological regulation database is set to 0.15. Therefore, the microecological swing threshold Sth is preferably 0.15.

[0121] When the microecological swing coefficient SHB is greater than 0.15, it indicates that the decline and duration of the current comprehensive microecological improvement state after reaching its maximum value have exceeded the upper quartile level of the historical effective control sample, and the microecological swing result is deemed unqualified; when the microecological swing coefficient SHB is less than or equal to 0.15, it indicates that the degree of maintenance of the comprehensive microecological state after reaching the maximum improvement state is within the swing range allowed by the historical effective control sample, and the microecological swing result is deemed qualified.

[0122] The calibration target for the microbial community-host response consistency threshold Jth is to determine a critical value for the microbial community-host response consistency coefficient that can distinguish between "microbial community-host response consistency reaching the standard state" and "microbial community-host response consistency not reaching the standard state".

[0123] This invention utilizes a historical database of effective microbial ecosystem regulation, prioritizing historical regulation samples that demonstrate satisfactory microbial community synergistic recovery and qualified microbial ecosystem reversion results. It obtains the comprehensive microbial ecosystem improvement and host improvement at each sampling time, and extracts the comprehensive microbial ecosystem improvement at the last sampling time, the host improvement at the last sampling time, the moment when the comprehensive microbial ecosystem improvement first enters the corresponding reference state interval, and the moment when the host improvement first enters the corresponding reference state interval. Simultaneously, Pearson correlation analysis is used to obtain the correlation coefficient between the comprehensive microbial ecosystem improvement and the host improvement. This invention uses the temporal relationship between the last state and the first entry into the reference state, along with temporal correlation, to evaluate the consistency of microbial community and host response.

[0124] According to the calculation method of the microbial community-host response consistency coefficient JSX, the historical microbial community-host response consistency coefficients corresponding to each historical regulation sample are obtained respectively. All historical microbial community-host response consistency coefficients are arranged in ascending order. The historical microbial community-host response consistency coefficient corresponding to the 25th position in the sorting result is taken as the lower quartile consistency response value, and the lower quartile consistency response value is determined as the microbial community-host response consistency threshold Jth.

[0125] The microbial community-host response consistency coefficient JSX, which considers the proximity between the last comprehensive microbial ecological improvement and the last host improvement, the time difference between their first entry into the corresponding reference state interval, and the correlation of the changing trends within the entire regulatory analysis time window, has its three components all limited to the range of 0 to 1. Therefore, the microbial community-host response consistency coefficient is generally within the range of 0 to 1. The larger the value, the more consistent the microbial community response and the host response are in terms of improvement degree, response time, and changing trend. Preferably, the lower quartile consistency response value of the historical microbial community-host response consistency coefficient in the historical effective microbial ecological regulation database is set to 0.50. Therefore, the microbial community-host response consistency threshold Jth is preferably 0.50.

[0126] When the consistency coefficient (JSX) between the microbial community and the host response is less than 0.50, it indicates that the overall consistency between the current comprehensive microecological improvement status and the host improvement status in terms of the degree of the last improvement, the time relationship to entering the reference state, or the trend of change has not reached the lower quartile level of historical effective regulatory samples, and it is determined that the consistency between the microbial community and the host response has not met the standard. When the consistency coefficient (JSX) between the microbial community and the host response reaches or exceeds 0.50, it indicates that the overall consistency between the current microbial community response and the host response has reached the lower quartile level of historical effective regulatory samples, and it is determined that the consistency between the microbial community and the host response has met the standard.

[0127] This invention further categorizes the effects of microbial regulation into different outcomes—insufficient regulation, post-improvement relapse, inconsistent microbial and host responses, and sustained effective regulation—by comprehensively considering the synergistic recovery state of the microbial community, the post-improvement relapse state, and the consistency of microbial and host responses. For each outcome, the invention re-evaluates the effect by maintaining the current dosage, increasing the candidate dosage, or extending the regulation process. This allows for different subsequent treatments based on the formation process and differences in host responses for the same final microbial community improvement result. Simultaneously, iterative updates are performed by re-acquiring microbial community and host response data in the next regulation cycle, ensuring that the evaluation of the regulation effect can continue to participate in subsequent regulation processes, improving the matching degree between subsequent candidate dosage adjustments and the current actual regulation state.

[0128] Please see Figure 2 A micro-ecological regulation effect analysis system based on artificial intelligence, specifically including:

[0129] The microecological data acquisition module is used to monitor research subjects that use probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts to regulate the intestinal microecology. It collects basic data of the regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed. Based on the historical regulation sample data, it identifies the microbial communities associated with improvement and the microbial communities associated with risk.

[0130] The microecological response modeling module is used to establish an artificial intelligence model training sample set based on historical regulation sample data. It uses a long short-term memory neural network for training to establish an artificial intelligence microecological response model. It also obtains the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators. Furthermore, it determines the reference state intervals for improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement.

[0131] The microbial community synergistic recovery analysis module is used to obtain the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time based on the microbial community data of the sample to be analyzed, improved associated microbial communities, risk associated microbial communities, effective recovery reference relative abundance of improved associated microbial communities, and effective inhibition reference relative abundance of risk associated microbial communities. Based on the recovery amount of improved associated microbial communities, the inhibition amount of risk associated microbial communities, the reference state interval of improved associated microbial communities, the reference state interval of risk associated microbial communities, and the basic data of the regulation process at each sampling time, the microbial community synergistic recovery coefficient is obtained.

[0132] The microecological swing analysis module is used to obtain the comprehensive microecological improvement at each sampling time based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, and to perform time-series change analysis on the comprehensive microecological improvement at each sampling time to obtain the microecological swing coefficient.

[0133] The microbiota-host response consistency analysis module is used to obtain the amount of inflammation improvement and mucosal barrier improvement at each sampling time based on the host response data, effective improvement reference values ​​of inflammatory indicators, and effective improvement reference values ​​of mucosal barrier indicators of the sample to be analyzed. It also obtains the amount of host improvement at each sampling time based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time. Based on the comprehensive microecological improvement, host improvement, comprehensive microecological improvement reference state interval, host improvement reference state interval, and basic data of the regulatory process at each sampling time, the microbiota-host response consistency coefficient is obtained.

[0134] The regulation effect evaluation and update module is used to conduct a comprehensive comparative analysis based on the microbial community synergistic recovery coefficient, the microecological swing coefficient, and the microbial community-host response consistency coefficient to obtain the microecological regulation effect evaluation results; and generate candidate dosage adjustment strategies for the next regulation cycle, and perform iterative updates of the regulation effect for the next regulation cycle.

[0135] This invention organizes data acquisition, artificial intelligence response modeling, microbial community collaborative recovery analysis, microbial community reversal analysis, microbial community and host response consistency analysis, and regulation effect evaluation and update according to the data dependencies between them. This allows the data and analysis results generated by the previous module to serve as the basis for the processing of the next module, thereby realizing a continuous processing flow from microbial regulation data acquisition to regulation effect evaluation and update for the next regulation cycle. This improves the convenience of the method in computer environment for process-oriented execution and repeated application.

[0136] It should be noted that all calculation formulas in this specification employ regression analysis, including but not limited to machine learning algorithms, to deeply analyze the collected relevant parameters and identify their natural trends and interrelationships. Specialized software, such as Python's Scikit-learn library or the R language, is used to automatically generate mathematical models that match the data. Then, cross-validation and other methods are used to objectively evaluate the model performance, and continuous feedback and optimization are combined to ensure that the created formulas truly reflect the inherent laws of the data, thereby guaranteeing their effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula undergo dimensionless processing within a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless processing techniques include, but are not limited to, min-max-normalization and Z-score standardization.

[0137] The algorithm of this invention is implemented as a Python script. Before executing the core logic, the program first executes a data loading module (e.g., using the widely used pandas library in Python) configured to read the aforementioned spreadsheet file and load its contents into the program's working memory (e.g., a DataFrame data structure). Subsequent algorithm steps will directly query and retrieve the required configuration parameters from this in-memory data structure.

[0138] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for analyzing the effects of microecological regulation based on artificial intelligence, characterized in that, The specific steps include: Step 1: Monitor the intestinal microecological regulation of the research subjects using probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts, collect basic data of the regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed, and determine the improvement-related microbial community and risk-related microbial community based on the historical regulation sample data; Step 2: Based on historical regulation sample data, establish an artificial intelligence model training sample set, use a long short-term memory neural network for training, and establish an artificial intelligence microecological response model; obtain the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators, and determine the reference state intervals for improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement. Step 3: Based on the microbial community data of the sample to be analyzed, the improved associated microbial community, the risk associated microbial community, the effective recovery reference relative abundance of the improved associated microbial community, and the effective inhibition reference relative abundance of the risk associated microbial community, obtain the recovery amount of the improved associated microbial community and the inhibition amount of the risk associated microbial community at each sampling time; based on the recovery amount of the improved associated microbial community, the inhibition amount of the risk associated microbial community, the reference state interval of the improved associated microbial community, the reference state interval of the risk associated microbial community, and the basic data of the regulation process at each sampling time, obtain the microbial community co-recovery coefficient; Step 4: Based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, obtain the comprehensive microecological improvement amount at each sampling time, and conduct a time-series change analysis on the comprehensive microecological improvement amount at each sampling time to obtain the microecological swing coefficient. Step 5: Based on the host response data, effective improvement reference values ​​of inflammatory indicators, and effective improvement reference values ​​of mucosal barrier indicators of the samples to be analyzed, obtain the amount of inflammation improvement and mucosal barrier improvement at each sampling time, and obtain the amount of host improvement at each sampling time based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time; based on the comprehensive microecological improvement, host improvement, comprehensive microecological improvement reference state interval, host improvement reference state interval, and basic data of the regulatory process at each sampling time, obtain the consistency coefficient between the microbiota and the host response. Step 6: Based on the synergistic recovery coefficient of the microbial community, the microecological swing coefficient, and the consistency coefficient between the microbial community and the host response, a comprehensive comparative analysis is conducted to obtain the evaluation results of the microecological regulation effect; and candidate dosage adjustment strategies for the next regulation cycle are generated, and the regulation effect of the next regulation cycle is iteratively updated.

2. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 1, characterized in that: Step one includes: S11. Monitor the regulation process of intestinal microecology in research subjects using probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts, and collect basic data of the regulation process, including: type of microecological regulation measure, current dose, start time of regulation, sampling time after regulation and total duration of regulation analysis time window; S12. Retrieve data from historical microecological regulation processes that are the same type as the current microecological regulation measures, collect historical regulation sample data, including: historical microbial community time-series data, historical inflammatory response time-series data, historical mucosal barrier response time-series data, and historical actual regulation effect results, and determine the improvement-related microbial communities and risk-related microbial communities based on the historical microbial community time-series data and historical actual regulation effect results. S13. Periodically monitor the gut microbiota of the research subjects and collect microbiota data of the samples to be analyzed, including: relative abundance of improved associated microbiota before regulation, relative abundance of improved associated microbiota at each sampling time after regulation, relative abundance of risk associated microbiota before regulation, and relative abundance of risk associated microbiota at each sampling time after regulation. S14. Periodically monitor the host response status of the research subjects and collect host response data of the samples to be analyzed, including: inflammatory marker detection values ​​before regulation, inflammatory marker detection values ​​at each sampling time after regulation, mucosal barrier marker detection values ​​before regulation, and mucosal barrier marker detection values ​​at each sampling time after regulation.

3. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 2, characterized in that: Step two includes: S21. Based on historical regulation sample data, the historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data corresponding to each historical regulation sample are arranged sequentially according to the sampling time before and after microecological regulation, according to the sample number. The historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data are standardized using the Z-score standardization method, and the mean and standard deviation of each type of data after Z-score standardization are recorded as Z-score standardization parameters. The standardized historical microbial community time-series data, historical inflammatory response time-series data, and historical mucosal barrier response time-series data are used as time-series input data to establish an artificial intelligence model training sample set. S22. Using a stratified random sampling method, the training sample set of the artificial intelligence model is divided into a training set and a validation set. Then, a long short-term memory neural network is used for training. The relative abundance of the improved associated bacteria, the relative abundance of the risk associated bacteria, the detection values ​​of inflammatory indicators and the detection values ​​of mucosal barrier indicators at the continuous sampling time are used as time-series inputs. The relative abundance of the improved associated bacteria, the relative abundance of the risk associated bacteria, the detection values ​​of inflammatory indicators and the detection values ​​of mucosal barrier indicators at the next sampling time are used as training targets. Backpropagation and gradient descent methods are used to update the network parameters. The network parameters are determined based on the prediction error of the validation set to establish an artificial intelligence microecological response model. S23. Using a label matching screening method, historical control samples corresponding to the effective control categories are selected based on the historical actual control effect results. Z-score standardization parameters are used to standardize the relative abundance of improved associated bacteria, relative abundance of risk-associated bacteria, inflammatory marker detection values, and mucosal barrier marker detection values ​​of each effective historical control sample at each sampling time before and after microecological control. The standardized data is then input into the artificial intelligence microecological response model to obtain the predicted sequences corresponding to each effective historical control sample. Z-score inverse standardization is used to restore each predicted sequence to its corresponding original data scale based on the Z-score standardization parameters. The prediction results at the last sampling time of each predicted sequence are extracted, and the average value calculation method is used to obtain the effective recovery reference relative abundance of improved associated bacteria, the effective inhibition reference relative abundance of risk-associated bacteria, the effective improvement reference detection value of inflammatory markers, and the effective improvement reference detection value of mucosal barrier markers, respectively. S24. Based on the predicted sequences restored to the original data scale corresponding to each effective historical regulation sample, as well as the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators, the degree of recovery of improved associated microbiota, the degree of inhibition of risk associated microbiota, the degree of improvement of comprehensive microecology, and the degree of improvement of host at each sampling time are normalized. The quartile statistical method is used to determine the reference state intervals of improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement, respectively.

4. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 3, characterized in that: Step three includes: S31. Based on the relative abundance of the improved associated microbial community before regulation, the relative abundance of the improved associated microbial community at each sampling time after regulation, and the effective recovery reference relative abundance of the improved associated microbial community, the normalized boundary constraint method is used to calculate the microbial community recovery ratio of each improved associated microbial community at each sampling time. S32. Based on the recovery ratio of each improved associated microbial community at each sampling time, the recovery amount of the improved associated microbial community at each sampling time is obtained by using the arithmetic mean method. S33. Based on the relative abundance of risk-associated microbial communities before regulation, the relative abundance of risk-associated microbial communities at each sampling time after regulation, and the effective suppression reference relative abundance of risk-associated microbial communities, the normalized boundary constraint method is used to obtain the microbial community suppression ratio of each risk-associated microbial community at each sampling time. S34. Based on the inhibition ratio of each risk-associated bacterial community at each sampling time, the inhibition amount of the risk-associated bacterial community at each sampling time is obtained by using the arithmetic mean method.

5. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 4, characterized in that: Step three also includes: S35. Based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, the time-series end-value extraction method is used to extract the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities corresponding to the last sampling time, and the last recovery amount of improved associated microbial communities and the last inhibition amount of risk associated microbial communities are obtained respectively; the interval matching and sequential retrieval method is used to perform interval matching according to the sampling time sequence, and the time when improved associated microbial communities first enter the corresponding reference state interval and the time when risk associated microbial communities first enter the corresponding reference state interval are obtained respectively; S36. Based on the last recovery amount of the improved associated microbial community, the last inhibition amount of the risk associated microbial community, the time when the improved associated microbial community first entered the corresponding reference state interval, the time when the risk associated microbial community first entered the corresponding reference state interval, and the total duration of the regulation analysis time window, calculate and obtain the microbial community co-recovery coefficient.

6. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 5, characterized in that: Step four includes: S41. Based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, the arithmetic mean method is used to obtain the comprehensive microecological improvement amount at each sampling time. S42. Based on the comprehensive microecological improvement at each sampling time, extreme value retrieval and time-series difference statistics are used to obtain the maximum comprehensive microecological improvement, the corresponding maximum improvement time, the comprehensive microecological improvement at the last sampling time, the number of sampling intervals where the comprehensive microecological improvement decreased after the maximum improvement time, and the total number of sampling intervals after the maximum improvement time. S43. Based on the maximum comprehensive microecological improvement, the comprehensive microecological improvement at the last sampling time, the number of sampling intervals where the comprehensive microecological improvement decreased after the maximum improvement time, and the total number of sampling intervals after the maximum improvement time, calculate and obtain the microecological swing coefficient.

7. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 6, characterized in that: Step five includes: S51. Based on the pre-modulation inflammatory marker detection values, the post-modulation inflammatory marker detection values ​​at each sampling time, and the effective improvement reference values ​​of inflammatory markers, the reference distance normalization and boundary constraint methods are used to obtain the inflammatory marker improvement ratio at each sampling time for each inflammatory marker. S52. Based on the improvement ratio of each inflammatory marker at each sampling time, the amount of inflammation improvement at each sampling time is obtained by using the arithmetic mean method. S53. Based on the mucosal barrier index detection values ​​before regulation, the mucosal barrier index detection values ​​at each sampling time after regulation, and the effective improvement reference detection values ​​of the mucosal barrier index, the reference distance normalization and boundary constraint methods are used to obtain the mucosal barrier index improvement ratio at each sampling time for each mucosal barrier index. S54. Based on the improvement ratio of each mucosal barrier index at each sampling time, the arithmetic mean method is used to obtain the amount of mucosal barrier improvement at each sampling time.

8. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 7, characterized in that: Step five also includes: S55. Based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time, the host improvement at each sampling time is obtained by using the arithmetic mean method. S56. Based on the host improvement amount at each sampling time, the host improvement amount at the last sampling time is obtained by using the time-series end value extraction method; based on the comprehensive microecological improvement amount and the host improvement amount at each sampling time, the correlation coefficient between the two is obtained by using the Pearson correlation analysis method; the interval matching and sequential retrieval methods are used to obtain the time when the comprehensive microecological improvement amount first enters the corresponding reference state interval and the time when the host improvement amount first enters the corresponding reference state interval, respectively. S57. Based on the comprehensive microecological improvement at the last sampling time, the host improvement at the last sampling time, the time when the comprehensive microecological improvement first enters the corresponding reference state interval, the time when the host improvement first enters the corresponding reference state interval, the correlation coefficient between the comprehensive microecological improvement and the host improvement, and the total duration of the regulation analysis time window, calculate and obtain the consistency coefficient between the microbial community and the host response.

9. The method for analyzing the effect of micro-ecological regulation based on artificial intelligence according to claim 8, characterized in that: Step six includes: S61. By setting preset thresholds for microbial community synergistic recovery, microecological reversal, and microbial community-host response consistency, and conducting a comprehensive comparative analysis with these thresholds and the corresponding coefficients, the microecological reversal coefficients and microbial community-host response consistency coefficients, the evaluation results of the microecological regulation effect are obtained, including: When the microbial community co-recovery coefficient is less than the microbial community co-recovery threshold, the sample to be analyzed is determined to be in a state of insufficient regulation. When the microbial community synergistic recovery coefficient is greater than or equal to the microbial community synergistic recovery threshold, and the microecological swing coefficient is greater than the microecological swing threshold, the sample to be analyzed is determined to be in the improved swing state. When the microbial community co-recovery coefficient is greater than or equal to the microbial community co-recovery threshold, the microecological swing coefficient is less than or equal to the microecological swing threshold, and the microbial community-host response consistency coefficient is less than the microbial community-host response consistency threshold, the sample to be analyzed is determined to be in a state of inconsistency between microbial community and host response. When the microbial community synergistic recovery coefficient is greater than or equal to the microbial community synergistic recovery threshold, the microecological swing coefficient is less than or equal to the microecological swing threshold, and the microbial community-host response consistency coefficient is greater than or equal to the microbial community-host response consistency threshold, the sample to be analyzed is determined to be in a state of continuous and effective regulation. S62. Based on the evaluation results of the microecological regulation effect and the current administration dose, generate candidate administration dose adjustment strategies for the next regulation cycle, including: When the evaluation result of the microecological regulation effect is a state of continuous and effective regulation, the current administration dose is maintained and the current administration dose is used as the candidate administration dose for the next regulation cycle; When the evaluation result of the microecological regulation effect is that the state has rebounded after improvement, the dose is increased by 5% to 10% based on the current dose, and the adjusted dose is used as the candidate dose for the next regulation cycle. When the evaluation result of the microecological regulation effect is that the regulation is insufficient, the dosage is increased by 10% to 15% based on the current dosage, and the adjusted dosage is used as the candidate dosage for the next regulation cycle. When the evaluation result of the microecological regulation effect is that the microbial community and the host response are inconsistent, the current administration dose is maintained, the current microecological regulation process is extended to the next sampling time, and the consistency coefficient between the microbial community and the host response is recalculated after acquiring the microbial community data and the host response data of the newly added sample to be analyzed. The candidate administration dose for the next regulation cycle is determined based on the updated evaluation result of the microecological regulation effect. S63. Implement subsequent microecological regulation based on the candidate dose for the next regulation cycle, and after obtaining the microbial community data and host response data of the sample to be analyzed corresponding to the next regulation cycle, recalculate and iteratively update the evaluation results of the microecological regulation effect and the candidate dose for the next regulation cycle.

10. A microecological regulation effect analysis system based on artificial intelligence, applied to the microecological regulation effect analysis method based on artificial intelligence as described in any one of claims 1 to 9, characterized in that, include: The microecological data acquisition module is used to monitor research subjects that use probiotics, prebiotics, dietary active ingredients or traditional Chinese medicine extracts to regulate the intestinal microecology. It collects basic data of the regulation process, historical regulation sample data, microbial community data of the sample to be analyzed, and host response data of the sample to be analyzed. Based on the historical regulation sample data, it identifies the microbial communities associated with improvement and the microbial communities associated with risk. The microecological response modeling module is used to establish an artificial intelligence model training sample set based on historical regulation sample data. It uses a long short-term memory neural network for training to establish an artificial intelligence microecological response model. It also obtains the effective recovery reference relative abundance of improved associated microbiota, the effective inhibition reference relative abundance of risk associated microbiota, the effective improvement reference detection value of inflammatory indicators, and the effective improvement reference detection value of mucosal barrier indicators. Furthermore, it determines the reference state intervals for improved associated microbiota, risk associated microbiota, comprehensive microecological improvement, and host improvement. The microbial community synergistic recovery analysis module is used to obtain the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time based on the microbial community data of the sample to be analyzed, improved associated microbial communities, risk associated microbial communities, effective recovery reference relative abundance of improved associated microbial communities, and effective inhibition reference relative abundance of risk associated microbial communities. Based on the recovery amount of improved associated microbial communities, the inhibition amount of risk associated microbial communities, the reference state interval of improved associated microbial communities, the reference state interval of risk associated microbial communities, and the basic data of the regulation process at each sampling time, the microbial community synergistic recovery coefficient is obtained. The microecological swing analysis module is used to obtain the comprehensive microecological improvement at each sampling time based on the recovery amount of improved associated microbial communities and the inhibition amount of risk associated microbial communities at each sampling time, and to perform time-series change analysis on the comprehensive microecological improvement at each sampling time to obtain the microecological swing coefficient. The microbiota-host response consistency analysis module is used to obtain the amount of inflammation improvement and mucosal barrier improvement at each sampling time based on the host response data, effective improvement reference values ​​of inflammatory indicators, and effective improvement reference values ​​of mucosal barrier indicators of the sample to be analyzed. It also obtains the amount of host improvement at each sampling time based on the amount of inflammation improvement and mucosal barrier improvement at each sampling time. Based on the comprehensive microecological improvement, host improvement, comprehensive microecological improvement reference state interval, host improvement reference state interval, and basic data of the regulatory process at each sampling time, the microbiota-host response consistency coefficient is obtained. The regulation effect evaluation and update module is used to conduct a comprehensive comparative analysis based on the microbial community synergistic recovery coefficient, the microecological swing coefficient, and the microbial community-host response consistency coefficient to obtain the microecological regulation effect evaluation results. It generates candidate dose adjustment strategies for the next regulation cycle and iteratively updates the regulation effect for the next regulation cycle.

Images