Method for identifying end point of black tea fermentation based on changes in key metabolites

Abstract

This application discloses a method for identifying the fermentation endpoint of black tea based on changes in key metabolites, relating to the field of tea processing technology. The method includes: acquiring sensory, metabolomic, and transcriptomic multi-source flavor characteristic data of a black tea sample under a preset fermentation time series; constructing a gene-metabolism-sensory multidimensional association network through association analysis, and screening components whose correlation coefficients with bitterness response values ​​and floral / fruity aroma attribute scores both exceed preset thresholds as core differential molecular markers; calculating the binding affinity of the markers to sensory receptors to determine weighted evaluation coefficients; calculating the weighted conversion rate of bitterness and the weighted cumulative intensity of aroma accordingly; and determining the corresponding fermentation time as the optimal fermentation endpoint when the rate of change in conversion rate is less than the threshold and the cumulative intensity is greater than or equal to the maximum peak value. This application can determine the optimal fermentation endpoint where bitterness and floral / fruity aromas reach a balance during black tea fermentation, ensuring the optimization and consistency of black tea flavor.

Description

Technical Field

[0001] This application relates to the field of tea processing technology, and in particular to a method for identifying the fermentation endpoint of black tea based on changes in key metabolites. Background Technology

[0002] Black tea processing typically includes withering, rolling, fermentation, and drying, with fermentation being the crucial stage determining the quality of black tea. During this process, polyphenol oxidases and peroxidases catalyze the oxidation and transformation of substances such as catechins, generating theaflavins, thearubigins, and theabrownins, thus affecting the color and taste of the tea liquor. Simultaneously, precursor substances such as carotenoids, amino acids, and glycoside-bound volatiles continuously degrade and transform, forming key aroma compounds such as floral and fruity aromas. As black tea production scales up, the demand for product quality stability and standardized control becomes increasingly prominent. Since fermentation time has a decisive impact on the transformation of bitter substances and the formation of floral and fruity aromas, there is an urgent need to establish a method that can scientifically identify the optimal fermentation endpoint for black tea.

[0003] Currently, in black tea production, determining the fermentation endpoint primarily relies on tea producers' experience based on sensory cues such as leaf color, aroma, and leaf quality, combined with control over processing conditions such as temperature, humidity, and time. Meanwhile, existing research generally recognizes a close relationship between fermentation duration and the flavor and quality of black tea.

[0004] However, existing practices still have significant shortcomings. On the one hand, relying solely on human experience and lacking precise biochemical or physical indicators easily leads to large batch-to-batch variations and poor repeatability, making it difficult to establish unified and stable standardized control specifications. On the other hand, although existing research has revealed the correlation between fermentation time and flavor quality, most of it remains at the descriptive level, lacking rigorous scientific criteria. A framework has not yet been established to systematically link the dynamic changes of key metabolites, the reduction of bitterness and astringency, the enhancement of floral and fruity aromas, and the evolution of sensory attributes during fermentation. Therefore, it is difficult to accurately determine the optimal fermentation endpoint and to provide directly applicable endpoint identification criteria for industrial production. Thus, determining the optimal fermentation endpoint that balances bitterness and astringency with floral and fruity aromas during black tea fermentation is an urgent problem to be solved. Summary of the Invention

[0005] The purpose of this application is to provide a method for identifying the fermentation endpoint of black tea based on changes in key metabolites, aiming to solve the technical problem of how to determine the optimal fermentation endpoint in which bitterness and floral and fruity aromas are balanced during the fermentation process of black tea.

[0006] To achieve the above objectives, this application proposes a method for identifying the fermentation endpoint of black tea based on changes in key metabolites, the method comprising: Acquire multi-source flavor characteristic data of the black tea sample to be tested under a preset fermentation time series. The multi-source flavor characteristic data includes sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic gene characteristics. Correlation analysis was performed on the multi-source flavor characteristic data to construct a gene-metabolism-sensory multidimensional correlation network characterizing the evolution of black tea quality; Based on the gene-metabolism-sensory multidimensional association network, metabolic components with a correlation coefficient greater than the first preset correlation threshold and a correlation coefficient greater than the second preset correlation threshold with the floral and fruity aroma attribute score were screened to obtain core differential molecular markers. The binding affinity of the core differential molecular markers to the target sensory receptors is calculated to obtain the sensory weight evaluation coefficient; Based on the sensory weight evaluation coefficients, the weighted conversion rate of bitter components and the weighted cumulative intensity of aroma components in the core differential molecular markers were calculated respectively. When the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, the corresponding fermentation time is taken as the optimal fermentation endpoint.

[0007] In one embodiment, the step of performing correlation analysis on the multi-source flavor characteristic data to construct a gene-metabolism-sensory multidimensional correlation network characterizing the evolution of black tea quality includes: The transcriptome gene features were modularly clustered using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules. Calculate the Pearson correlation coefficient between the characteristic gene value of each co-expression module and each sensory attribute in the sensory digital features, and the content of each metabolite in the metabolomic chemical features; Target modules whose Pearson correlation coefficient is greater than a preset correlation threshold are selected; Hub genes with a preset connectivity are extracted from the target module, and differential metabolites in the metabolomic chemical characteristics corresponding to the target module are identified. Based on the correlation calculation results among the hub gene, the differential metabolites, and the sensory attributes in the sensory digitization features, a topological connection relationship among the three is established. Gene-metabolism-sensory multidimensional network is generated based on the topological connections.

[0008] In one embodiment, the step of modularly clustering the transcriptome gene features using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules includes: Based on the expression levels of each gene in the transcriptome gene characteristics, calculate the Pearson correlation coefficient matrix between genes; The Pearson correlation coefficient matrix is ​​exponentially operated on each element according to a preset soft threshold to obtain an adjacency matrix that conforms to the scale-free network distribution. The topological overlap measure between genes is calculated based on the adjacency matrix to obtain the topological overlap matrix used to characterize the connectivity between genes; Using the heterogeneity of the topological overlap matrix as a distance metric, hierarchical clustering is performed to obtain a gene clustering dendrogram; The gene clustering dendrogram is branched using a dynamic tree pruning algorithm, and similar modules are merged based on the correlation of characteristic genes in each branch to obtain multiple co-expression modules.

[0009] In one embodiment, the step of acquiring multi-source flavor characteristic data of the black tea sample under a preset fermentation time series, wherein the multi-source flavor characteristic data includes sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic genetic characteristics, includes: The taste response potential values ​​of the black tea sample to be tested are collected by a taste sensor to obtain a taste feature vector containing bitter, astringent, fresh and sweet response values. The response signal intensity of the black tea sample under test in the headspace environment is collected by a gas-sensitive sensor array, and the response signal intensity is matrixed according to the sensor channel dimension to obtain the odor feature vector. The taste feature vector and the odor feature vector are concatenated and fused to obtain sensory digital features; The non-volatile and volatile compounds in the black tea sample were qualitatively and quantitatively analyzed by liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry, respectively, to obtain the metabolomic chemical characteristics. Transcript abundance information of the black tea sample under test at various time points was obtained by high-throughput sequencing to obtain transcriptome gene characteristics. The sensory digital features, metabolomic chemical features, and transcriptomic gene features are aligned along a timeline according to a preset fermentation time sequence to obtain multi-source flavor feature data.

[0010] In one embodiment, the step of screening metabolic components with correlation coefficients greater than a first preset correlation threshold and correlation coefficients greater than a second preset correlation threshold with floral and fruity aroma attribute scores based on the gene-metabolism-sensory multidimensional association network to obtain core differential molecular markers includes: Bitter taste components with a correlation coefficient greater than a first preset correlation threshold are extracted from the gene-metabolism-sensory multidimensional association network. The bitter taste components include epigallocatechin gallate. Aroma components with correlation coefficients greater than a second preset correlation threshold with floral and fruity aroma attribute scores are extracted from the gene-metabolism-sensory multidimensional association network. The aroma components include geraniol, damastone, and phenylacetaldehyde. Amino acid components that are positively correlated with the umami and sweetness response value were extracted from the gene-metabolism-sensory multidimensional association network, and the amino acid components included alanyl-glutamine. From the gene-metabolism-sensory multidimensional association network, pivot genes that have co-expression relationships with the bitter taste component, the aroma component and the amino acid component are screened, and the pivot genes include deubiquitinase genes; The core differential molecular markers were obtained by summarizing the bitter taste components, the aroma components, the amino acid components, and the hub gene.

[0011] In one embodiment, the step of calculating the binding affinity of the core differential molecular marker to the target sensory receptor to obtain the sensory weight evaluation coefficient includes: The three-dimensional chemical structures of each compound molecule in the core differential molecular markers were obtained as docking ligands; Obtain structural models of receptor proteins, including bitter taste receptors, sweet taste receptors, and olfactory receptors. The docking ligand is flexibly docked with the receptor protein structure model, and the binding free energy is calculated. The sensory intensity contribution of each compound molecule is scored based on the magnitude of its binding free energy, resulting in a sensory weight evaluation coefficient.

[0012] In one embodiment, the step of calculating the weighted conversion rate of the bitter taste component and the weighted cumulative intensity of the aroma component in the core differential molecular markers based on the sensory weight evaluation coefficients includes: The real-time weighted bitterness value is obtained by weighting and summing the real-time concentrations of bitter components in the core differential molecular markers during fermentation based on the sensory weight evaluation coefficients. The weighted conversion rate is obtained by calculating the reduction ratio of the real-time weighted bitterness value relative to the initial stage of fermentation. The real-time concentrations of aroma components in the core differential molecular markers during fermentation are weighted according to the sensory weight evaluation coefficients to obtain real-time weighted aroma values. The weighted cumulative intensity is obtained by calculating the increase factor of the real-time weighted aroma value relative to the fermentation initiation stage.

[0013] Furthermore, to achieve the above objectives, this application also proposes a device for identifying the fermentation endpoint of black tea based on changes in key metabolites, the device comprising: The data acquisition module is used to acquire multi-source flavor feature data of the black tea sample to be tested under a preset fermentation time series. The multi-source flavor feature data includes sensory digital features, metabolomic chemical features and transcriptomic gene features. The association network construction module is used to perform association analysis on the multi-source flavor feature data and construct a gene-metabolism-sensory multidimensional association network characterizing the evolution of black tea quality. The marker screening module is used to screen metabolic components with a correlation coefficient greater than a first preset correlation threshold and a correlation coefficient greater than a second preset correlation threshold with the floral and fruity aroma attribute score based on the gene-metabolism-sensory multidimensional association network, thereby obtaining core differential molecular markers. The affinity calculation module is used to calculate the binding affinity between the core differential molecular marker and the target sensory receptor to obtain the sensory weight evaluation coefficient. The weight quantization module is used to calculate the weighted conversion rate of bitter taste components and the weighted cumulative intensity of aroma components in the core differential molecular markers based on the sensory weight evaluation coefficients. The endpoint determination module is used to determine the optimal fermentation endpoint when the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value.

[0014] Furthermore, to achieve the above objectives, this application also proposes a device for identifying the fermentation endpoint of black tea based on changes in key metabolites. The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites as described above.

[0015] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites as described above.

[0016] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites as described above.

[0017] One or more technical solutions proposed in this application have at least the following technical effects: First, multi-source flavor characteristic data of the black tea samples under a pre-defined fermentation time series were acquired, including sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic genetic characteristics, providing a data foundation for subsequent analysis. Next, correlation analysis was performed on these multi-source flavor characteristic data to construct a gene-metabolism-sensory multi-dimensional correlation network. By integrating information from different data levels, the complex mechanisms of black tea quality changes were revealed more comprehensively. Then, based on this network, metabolic components with correlation coefficients greater than a first pre-defined threshold for bitterness response and greater than a second pre-defined threshold for floral and fruity aroma attribute scores were screened to obtain core differential molecular markers, precisely locating key metabolites characterizing black tea flavor transformation. Next, the binding affinity of these core differential molecular markers to target sensory receptors was calculated, and sensory weight evaluation coefficients were obtained. This helps quantify the contribution of different metabolites to sensory experience, providing theoretical support for flavor optimization. Subsequently, based on the sensory weight evaluation coefficients, the weighted conversion rate of bitterness components and the weighted cumulative intensity of aroma components in the core differential molecular markers were calculated, helping to monitor changes in bitterness and aroma during fermentation and ensure that flavor components reach an ideal balance. Finally, when the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, the corresponding fermentation time is determined as the optimal fermentation endpoint. This application can determine the optimal fermentation endpoint in the fermentation process of black tea to achieve a balance between bitterness and floral and fruity aromas, ensuring the optimization and consistency of the flavor of black tea. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an embodiment of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites in this application. Figure 2 This is a flowchart illustrating Example 2 of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites in this application. Figure 3 This is a schematic diagram of the module structure of the black tea fermentation endpoint identification device based on changes in key metabolites according to an embodiment of this application. Figure 4This is a schematic diagram of the hardware operating environment involved in the black tea fermentation endpoint identification method based on key metabolite changes in the embodiments of this application.

[0021] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0023] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0024] It should be noted that the executing entity of this application embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or a black tea processing control system capable of realizing the above functions. The following description uses a black tea processing control system as an example to illustrate this embodiment and the subsequent embodiments.

[0025] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0026] Based on this, embodiments of this application provide a method for identifying the fermentation endpoint of black tea based on changes in key metabolites, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites according to this application.

[0027] In this embodiment, the method for identifying the fermentation endpoint of black tea based on changes in key metabolites includes steps S10 to S60: Step S10: Obtain multi-source flavor characteristic data of the black tea sample to be tested under a preset fermentation time series. The multi-source flavor characteristic data includes sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic gene characteristics. Step S20: Perform correlation analysis on the multi-source flavor feature data to construct a gene-metabolism-sensory multidimensional correlation network characterizing the evolution of black tea quality; Step S30: Based on the gene-metabolism-sensory multidimensional association network, screen metabolic components with a correlation coefficient greater than the first preset correlation threshold and a correlation coefficient greater than the second preset correlation threshold with the floral and fruity aroma attribute score to obtain core differential molecular markers; Step S40: Calculate the binding affinity between the core differential molecular marker and the target sensory receptor to obtain the sensory weight evaluation coefficient; Step S50: Based on the sensory weight evaluation coefficient, calculate the weighted conversion rate of the bitter taste component and the weighted cumulative intensity of the aroma component in the core differential molecular markers. Step S60: When the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, the corresponding fermentation time is taken as the optimal fermentation endpoint.

[0028] It should be noted that the black tea sample to be tested refers to a tea sample prepared under the same black tea processing conditions and used for flavor characteristic detection and fermentation endpoint determination. Preferably, it is a representative sample collected at different fermentation times, such as samples corresponding to fermentation times of 2.5 hours, 3.5 hours, 4.5 hours, 5.5 hours, and 6.5 hours. The preset fermentation time series refers to an ordered set of multiple sampling time points pre-set before fermentation begins, used to continuously characterize the dynamic changes of black tea during fermentation. These time points can be 2.5 hours, 3.5 hours, 4.5 hours, 5.5 hours, and 6.5 hours, and can also be adjusted according to process requirements. Sensory digital characteristics refer to the taste and aroma characteristics obtained through electronic tongue, electronic nose, and quantitative descriptive analysis, which can be numerically represented. These include bitterness, astringency, umami, and sweetness response values, as well as attribute scores for fruity, floral, sweet, woody, and grassy aromas. Metabolomics chemical characteristics refer to the chemical composition information obtained through metabolomics detection, including the types, relative contents, and trends of variation of non-volatile metabolites and volatile organic compounds, as well as the metabolite profile characteristics related to flavor formation.

[0029] Transcriptome gene characteristics refer to gene expression information obtained through transcriptome sequencing, including gene expression abundance at each fermentation stage, differentially expressed genes, co-expressed modules, and gene expression changes related to flavor-related metabolic pathways. Characterizing the evolution of black tea quality refers to the dynamic process reflecting the gradual change in taste and aroma from an initial state to a harmonious and balanced state during fermentation, primarily manifested as a decrease in bitterness and astringency, an increase in floral and fruity aromas, and a shift in the overall flavor profile from raw and astringent to mellow and harmonious. The gene-metabolism-sensory multidimensional association network is a network model constructed based on the statistical associations between gene expression data, metabolite data, and sensory evaluation data, used to reveal the correspondence and synergistic effects between gene regulation, metabolite transformation, and changes in sensory attributes. Bitterness and astringency response values ​​refer to the digital measurement results used to characterize the intensity of bitterness and astringency in black tea samples, typically represented by the bitterness response value, astringency response value, or a combined characterization value derived from both by an electronic tongue. The correlation coefficient is a statistical measure used to measure the direction and strength of the correlation between two variables; a value closer to 1 indicates a higher degree of positive correlation, while a value closer to 0 indicates a weaker correlation.

[0030] The first preset correlation threshold refers to a pre-set judgment boundary used to determine whether a certain metabolic component and the bitterness / astringency response value reach a significant positive correlation. In this embodiment, it is preferably 0.8. The floral and fruity aroma attribute score refers to the quantitative score result used to characterize the intensity of floral and fruity aromas in a black tea sample. It is usually derived from quantitative descriptive analysis or other sensory digital evaluation methods, and is used to reflect the high-quality aroma performance of the sample during fermentation. The second preset correlation threshold refers to a pre-set judgment boundary used to determine whether a certain metabolic component and the floral and fruity aroma attribute score reach a significant positive correlation. In this embodiment, it is preferably 0.8. Metabolic components refer to individual chemical components that can be detected by metabolomics during the fermentation of black tea, including specific metabolites such as catechins, amino acids, flavonoids, and volatile aroma compounds. Core differential molecular markers refer to key metabolites that undergo significant changes between different fermentation stages and are closely related to both the reduction of bitterness / astringency and the enhancement of floral and fruity aromas. They are used as molecular basis for identifying the flavor transformation of black tea and determining the optimal fermentation endpoint. Target sensory receptors refer to taste receptors or olfactory receptors that are related to the perception of key flavors in black tea and are used for molecular docking analysis.

[0031] Binding affinity refers to the ease and stability of binding between a core differential molecular marker and a target sensory receptor. It is usually expressed as the binding energy or a score obtained from molecular docking. The lower the binding energy, the more stable the binding and the stronger the interaction. Sensory weighting evaluation coefficients are weighting parameters calculated based on the binding affinity between core differential molecular markers and target sensory receptors. They characterize the relative contribution of different molecules to the perception of bitterness or floral / fruity aromas. Bitterness components refer to the metabolic components in core differential molecular markers that are mainly related to the formation of bitterness and / or astringency, typically including catechins and other compounds that are significantly positively correlated with increased bitterness. Weighted conversion rate is a comprehensive index obtained by quantifying the degree of conversion of bitterness components during fermentation and then weighting it with the corresponding sensory weighting evaluation coefficients. It reflects the overall efficiency of the conversion of bitterness-related substances from high to low levels. Aroma components refer to the metabolic components in core differential molecular markers that are mainly related to the formation of floral and / or fruity aromas. They typically include volatile aroma compounds such as geraniol, damastones, and phenylacetaldehyde, as well as other compounds related to the enhancement of floral and fruity aromas.

[0032] Weighted cumulative intensity refers to a comprehensive index obtained by quantifying the cumulative level or signal intensity of aroma components during fermentation and then weighting and summing them together with corresponding sensory evaluation coefficients. It reflects the overall degree of formation of floral and fruity aromas. The preset rate of change threshold is a pre-set boundary value used to determine whether the weighted conversion rate change tends to level off. When the rate of change of the weighted conversion rate is lower than this threshold, it indicates that the further transformation of bitterness-related substances has approached stability. The preset maximum peak value is a pre-set reference value used to determine whether the weighted cumulative intensity of aroma components has reached its peak state. When the weighted cumulative intensity reaches or exceeds this value, it indicates that the formation of floral and fruity aromas has reached the target high value range. The corresponding fermentation duration refers to the specific fermentation time point in the preset fermentation time series that corresponds to the above judgment conditions, i.e., the fermentation duration corresponding to the stabilization of bitterness transformation and the achievement of target floral and fruity aroma accumulation. The optimal fermentation endpoint refers to the point at which fermentation is stopped during the fermentation process of black tea, when bitterness and astringency are significantly reduced, floral and fruity aromas are fully formed, and the overall flavor is harmonious and balanced. It is used to guide precise fermentation control and quality stabilization in industrial production.

[0033] Understandably, the black tea processing control system first establishes a continuous monitoring process around the fermentation process of the black tea sample to be tested. It sequentially acquires corresponding samples at multiple fermentation stages according to a preset fermentation time sequence, and simultaneously acquires sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic genetic characteristics at each time point. These are then organized, matched, and stored in the same chronological order to form multi-source flavor characteristic data that reflects the entire fermentation process. This is done to avoid relying solely on a single time point or type of data, which fails to comprehensively reflect the flavor formation process of black tea. Next, the black tea processing control system performs joint analysis of the multi-source flavor characteristic data, comparing the direction and magnitude of changes in various data types at different time points, as well as their corresponding relationships. Sensory changes, chemical composition changes, and gene expression changes are placed within the same analytical framework, constructing a gene-metabolism-sensory multi-dimensional correlation network characterizing the evolution of black tea quality. This transforms the originally scattered data into an identifiable and traceable correlation structure. This approach aims to identify key information that truly correlates with changes in black tea quality, rather than simply comparing isolated data.

[0034] Next, the black tea processing control system, based on the gene-metabolism-sensory multidimensional association network, systematically retrieves metabolic components that are highly correlated with both bitterness and astringency response values ​​and floral / fruity aroma attribute scores. Metabolic components meeting these dual criteria are identified as core differential molecular markers. This allows subsequent analysis to focus on the group of key molecules most relevant to the fermentation endpoint, reducing irrelevant information interference and improving the accuracy of endpoint identification. Building on this, the black tea processing control system further calculates the binding affinity between the core differential molecular markers and target sensory receptors, and derives sensory weight evaluation coefficients based on the calculation results. This ensures that different core differential molecular markers are considered in subsequent judgments according to their magnitude of sensory impact, rather than being simply treated as equally important. This approach makes the analysis results more closely reflect the actual flavor formation effect.

[0035] Subsequently, the black tea processing control system, based on sensory weight evaluation coefficients, performs weighted calculations on the changes of bitter and astringent components and aroma components in the core differential molecular markers at different fermentation time points. This yields the weighted conversion rate of the bitter and astringent components and the weighted cumulative intensity of the aroma components. The former reflects whether the changes related to bitterness and astringency have stabilized, while the latter reflects whether the changes related to floral and fruity aromas have reached a high level. This is done to transform the complex flavor changes during fermentation into directly comparable and quantifiable results. Finally, the black tea processing control system evaluates the calculation results at each fermentation time point. When the weighted conversion rate at a certain time point is found to be less than a preset rate of change threshold, and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, it is determined that there is little room for further improvement in bitterness and astringency after that time point, while the floral and fruity aromas have reached the target level. Therefore, the fermentation time corresponding to that time point is determined as the optimal fermentation endpoint. This avoids premature termination of fermentation leading to insufficient aroma and also avoids flavor dulling or quality deviation caused by continued fermentation.

[0036] As an example, the step of obtaining multi-source flavor feature data of the black tea sample under a preset fermentation time series, wherein the multi-source flavor feature data includes sensory digital features, metabolomic chemical features, and transcriptomic gene features, includes: acquiring the taste response potential value of the black tea sample under test through a taste sensor to obtain a taste feature vector containing bitter, astringent, umami, and sweet response values; acquiring the response signal intensity of the black tea sample under test in a headspace environment through a gas-sensitive sensor array, and constructing the response signal intensity into a matrix according to the sensor channel dimension to obtain an odor feature vector; cascading and fusing the taste feature vector and the odor feature vector to obtain sensory digital features; qualitatively and quantitatively analyzing the non-volatile compounds and volatile compounds in the black tea sample under test through liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry to obtain metabolomic chemical features; acquiring the transcript abundance information of the black tea sample under test at each time point through high-throughput sequencing to obtain transcriptomic gene features; and aligning the sensory digital features, the metabolomic chemical features, and the transcriptomic gene features according to the preset fermentation time series to obtain multi-source flavor feature data.

[0037] It should be noted that a taste sensor is a sensor device used to perceive the taste attributes of a black tea sample and output electrical signals. It typically includes sensing units sensitive to different tastes such as bitterness, astringency, umami, and sweetness, converting taste differences that were originally perceived manually into measurable and comparable digital signals. The taste response potential value refers to the potential change generated when the taste sensor comes into contact with the black tea sample, characterizing the intensity of stimulation in the corresponding taste dimension, expressed as the relative potential difference between the sample measurement and a reference state. The taste feature vector is a set of values ​​formed by combining the response results of the black tea sample in multiple taste dimensions in a predetermined order, used to comprehensively characterize the taste profile of the sample; for example, it may be composed of response values ​​for bitterness, astringency, umami, and sweetness. A gas sensor array is a detection array composed of multiple gas sensing units with varying sensitivities to different volatile gases or odor components, used to simultaneously acquire aroma information from multiple response channels of the black tea sample to improve the ability to recognize complex odor profiles.

[0038] The headspace environment refers to the gas phase space above the black tea sample being tested, formed by the volatile components released from the sample. In aroma detection, it is typically formed through equilibrium within a sealed container and reflects the odor composition of the sample that can be captured by the gas-sensitive sensor. The response signal intensity refers to the signal magnitude generated by each sensing unit in the gas-sensitive sensor array upon contact with the headspace environment of the black tea sample. It reflects the degree of response of the corresponding channel to the volatile components of the sample and the level of aroma stimulation. The sensor channel dimension refers to the dimension in which data is differentiated and organized according to each independent sensing unit or detection channel in the gas-sensitive sensor array. Each channel corresponds to a type of response feature, used to maintain the independence and comparability of the output information of different sensing units. The odor feature vector is a set of values ​​formed by arranging the response signal intensities of the gas-sensitive sensor array on each sensor channel in a preset order. It is used to represent the overall odor profile and aroma response pattern of the black tea sample being tested. Sensory digital features refer to the feature data formed by integrating the taste feature vector and the odor feature vector, which can digitally characterize the sensory attributes of the black tea sample being tested. It is used to objectively reflect the comprehensive sensory performance of the sample at both the taste and aroma levels.

[0039] Liquid chromatography-mass spectrometry (LC-MS / MS) is an analytical technique that combines the separation capabilities of liquid chromatography with the detection capabilities of mass spectrometry. It is used to separate, identify, and quantify non-volatile compounds in black tea samples, thereby obtaining information on their non-volatile metabolic components. Gas chromatography-mass spectrometry (GC-MS / MS) is an analytical technique that combines the separation capabilities of gas chromatography with the detection capabilities of mass spectrometry. It is used to separate, identify, and quantify volatile compounds in black tea samples, thereby obtaining information on their volatile aroma components. Non-volatile compounds refer to chemical components that do not readily volatilize into the gas phase under conventional detection or processing conditions and are mainly present in the tea infusion or tea matrix. Examples include catechins, flavonoids, amino acids, and peptides, which are closely related to flavor formation. Volatile compounds refer to chemical components that can be released from the black tea sample into the gas phase and perceived as aroma, including esters, alcohols, terpenes, aldehydes, ethers, and other substances that contribute to floral, fruity, and other aroma properties.

[0040] Metabolomics chemical characteristics refer to the chemical compositional information obtained after qualitative and quantitative analysis of black tea samples using liquid chromatography-mass spectrometry (LC-MS) and gas chromatography-mass spectrometry (GC-MS). This includes the composition and content of both non-volatile and volatile compounds, reflecting the dynamic changes of metabolites during fermentation. High-throughput sequencing refers to sequencing technologies capable of parallel analysis of large numbers of nucleic acid molecules, used to rapidly obtain large-scale transcriptional information from black tea samples to support comparative gene expression analysis at different fermentation time points. Transcript abundance information refers to the expression level or relative expression level of each transcript in the black tea sample at a specific time point, reflecting the transcriptional activity and dynamic changes of the corresponding gene during fermentation. Transcriptome gene characteristics refer to gene expression feature information obtained based on high-throughput sequencing and formed after alignment, assembly, quantification, and analysis. This includes at least the expression abundance and changes of transcripts at different fermentation time points, characterizing the dynamic features of gene expression during fermentation.

[0041] Understandably, the black tea processing control system first takes samples of the black tea to be tested at each fermentation time point according to the preset fermentation time sequence, and performs taste detection on the sample corresponding to each time point. The taste sensor collects the taste response potential values ​​of the black tea sample to be tested in various dimensions such as bitterness, astringency, freshness and sweetness. Then, these response values ​​are arranged in a preset order to form the taste feature vector corresponding to that time point. This is done in order to transform the taste state that changes over time during the black tea fermentation process into continuous data that can be recorded and compared. Secondly, the black tea processing control system detects the odor of the black tea sample to be tested at the same time point. First, it acquires the odor signal of the black tea sample under the headspace environment. Then, the gas-sensitive sensor array collects the response signal intensity output by different sensor channels. The data of each channel is organized and matrixed according to the sensor channel dimension to form the odor feature vector corresponding to the time point. Then, the taste feature vector and the odor feature vector are cascaded and fused in sequence to obtain the sensory digital features of the black tea sample to be tested at that time point. This is done in order to unify the taste information and odor information into the same set of sensory data, so that subsequent analysis can simultaneously reflect the overall change of black tea flavor, rather than analyzing a single sensory signal in isolation.

[0042] Then, the black tea processing control system continued to conduct chemical component detection and gene expression detection on the black tea samples to be tested at the same time point. The non-volatile compounds and volatile compounds in the black tea samples to be tested were qualitatively and quantitatively processed by liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry, respectively, to obtain the metabolomic chemical characteristics corresponding to that time point. At the same time, the transcript abundance information of the black tea samples to be tested at that time point was obtained by high-throughput sequencing to obtain the transcriptomic gene characteristics. This is done in order to further capture the changes in components and expression during fermentation in addition to sensory changes, so that subsequent analysis can not only see the changes in flavor, but also correspond to the underlying internal changes. Finally, the black tea processing control system systematically matches, organizes, and aligns the sensory digital features, metabolomic chemical features, and transcriptomic genetic features corresponding to each fermentation time point according to the preset fermentation time sequence. This allows data from different sources at the same time point to form matching data units, which are then combined into a complete dataset according to the order of fermentation to obtain multi-source flavor feature data. This ensures that various types of data are strictly matched in time during subsequent analysis (e.g., sensory data, chemical data, and genetic data at the same fermentation stage are matched), thereby accurately reflecting the continuous process of flavor formation and evolution during black tea fermentation.

[0043] As an example, the step of screening metabolic components with correlation coefficients greater than a first preset correlation threshold and correlation coefficients greater than a second preset correlation threshold based on the gene-metabolism-sensory multidimensional association network to obtain core differential molecular markers includes: extracting bitter taste components from the gene-metabolism-sensory multidimensional association network that have correlation coefficients greater than a first preset correlation threshold and correlation coefficients greater than a second preset correlation threshold, wherein the bitter taste components include epigallocatechin gallate; and extracting metabolic components from the gene-metabolism-sensory multidimensional association network that have correlation coefficients greater than a first preset correlation threshold and correlation coefficients greater than a second preset correlation threshold and correlation coefficients greater than a second preset correlation threshold. Two aroma components with preset relevant thresholds, including geraniol, damastones, and phenylacetaldehyde; amino acid components positively correlated with the sweet taste response value are extracted from the gene-metabolism-sensory multidimensional association network, including alanyl-glutamine; hub genes co-expressed with the bitter taste component, aroma components, and amino acid components are screened from the gene-metabolism-sensory multidimensional association network, including deubiquitinase genes; and core differential molecular markers are obtained by summarizing the bitter taste component, aroma components, amino acid components, and hub genes.

[0044] It should be noted that amino acid components refer to amino acid metabolic components, used to characterize the molecular basis for enhancing umami and sweetness during black tea fermentation. These include alanine-glutamine, which is significantly associated with enhanced sweetness and umami and can serve as an important component in determining flavor optimization. Co-expression relationships refer to the corresponding, synergistic, or synchronous expression relationships between different molecular entities during fermentation. This means that the changing trends of related entities at multiple fermentation time points are consistent or significantly correlated. This relationship mainly indicates that bitterness components, aroma components, amino acid components, and related genes are not isolated changes but exist within the same regulatory network or functional framework. Hub genes are genes located at core connection positions in the co-expression network or key modules, possessing high node connectivity and playing a crucial regulatory role in flavor-related metabolic changes. These genes are not ordinary associated genes but key nodes connecting dynamic changes in the transcriptome with the accumulation of flavor metabolites. Therefore, they can serve as important regulatory factors and candidate molecular markers for explaining the quality formation mechanism during black tea fermentation. Deubiquitinase genes are genes that encode deubiquitinases. Their expression products can participate in the removal of protein ubiquitination modifications, thereby affecting related signaling pathways, transcriptional regulation processes, or metabolic regulation processes. Deubiquitinase genes include cysteine-type deubiquitinase genes, which can serve as pivotal genes related to the accumulation of specific aroma components and be used to explain the relationship between gene regulation and metabolic changes during flavor formation.

[0045] Understandably, firstly, the black tea processing control system extracts bitter and astringent components from a gene-metabolism-sensory multidimensional correlation network that have a correlation coefficient greater than a first preset correlation threshold with the bitter and astringent response value. Specifically, the system calculates the correlation between each metabolite in the multidimensional correlation network and bitterness and astringency, selecting metabolites with a significant correlation to the intensity of bitterness and astringency, such as epigallocatechin gallate. These bitter and astringent components have a significant impact on the taste formation of black tea and therefore require special attention. Secondly, the system extracts aroma components from the same network that have a correlation coefficient greater than a second preset correlation threshold with the floral and fruity aroma attribute score. By calculating the correlation between aroma components and floral and fruity aroma scores, key aroma molecules are extracted, such as geraniol, damastones, and phenylacetaldehyde. These aroma components play an important role in the flavor and aroma expression of black tea and therefore need to be screened.

[0046] Then, using similar analytical methods, the system extracted amino acid components positively correlated with the sweetness response value, specifically selecting amino acids such as alanine-glutamine, which significantly contribute to the formation of sweetness. Next, by screening co-expression relationships, the system further identified key genes co-expressed with bitterness components, aroma components, and amino acid components. Key genes, such as deubiquitinase genes, have high network connectivity and regulatory effects, influencing multiple metabolic pathways and flavor formation processes, and therefore needed to be included in the analysis. Finally, the system summarized the extracted bitterness components, aroma components, amino acid components, and key genes, and through multi-dimensional correlation analysis, obtained a set of core differential molecular markers. These markers can accurately indicate flavor changes during the fermentation process of black tea, helping to accurately determine the optimal fermentation endpoint, thereby improving the precision of black tea production control and quality consistency.

[0047] As an example, the step of calculating the binding affinity between the core differential molecular marker and the target sensory receptor to obtain the sensory weight evaluation coefficient includes: obtaining the three-dimensional chemical structure of each compound molecule in the core differential molecular marker as a docking ligand; obtaining the receptor protein structure model including bitter taste receptor, sweet taste receptor and olfactory receptor; flexibly docking the docking ligand with the receptor protein structure model and calculating the binding free energy; scoring the sensory intensity contribution of each compound molecule according to the magnitude of the binding free energy to obtain the sensory weight evaluation coefficient.

[0048] It should be noted that the three-dimensional chemical structure refers to the spatial configuration and atomic connection mode of each compound in the core differential molecular marker. It is usually calculated by computational chemistry software based on the molecular formula and interatomic bonding relationships, and is used to simulate the interaction between molecules and receptors. Docking ligands are molecules used to bind to receptor proteins during molecular docking. Here, they refer to compound molecules extracted from the core differential molecular marker as docking targets, and their binding affinity to the receptor is calculated. Bitter taste receptors are receptor proteins responsible for sensing bitterness, such as T2R14, which is one of the key receptors for human bitterness perception and participates in the recognition of bitter taste molecules. Sweet taste receptors are receptor proteins involved in sweet taste perception, such as T1R2 / T1R3, which are heterodimers of sweet taste receptors and trigger the perception of sweetness by binding to sweet taste molecules.

[0049] Olfactory receptors are receptor proteins that bind to odor molecules and transmit olfactory signals, such as Q9P1Q5, and are involved in the perception of certain aroma molecules. Receptor protein structural models are three-dimensional spatial structural models of receptor proteins constructed using bioinformatics methods, usually derived from experimental data or computational simulations, used to simulate the interaction between receptors and ligands. Binding free energy refers to the change in free energy required for ligands to bind to receptors during molecular docking, used to measure the binding affinity between ligands and receptors; the lower the binding free energy, the more stable the binding between the ligand and receptor, and the stronger the affinity. Sensory intensity contribution refers to the intensity of the sensory effect of a compound molecule after binding to a receptor, usually estimated by the magnitude of the binding free energy; molecules with lower binding free energies generally have a stronger effect in sensory perception.

[0050] Understandably, firstly, after obtaining the core differential molecular markers, the black tea processing control system retrieves the three-dimensional chemical structure data of each compound molecule and organizes the corresponding structures uniformly. This ensures that each compound molecule can enter the subsequent analysis process with a calculable spatial configuration, serving as the basis for simulating its interaction with the target sensory receptor. This is done to avoid judging its sensory effect solely based on the compound's name or content, but rather to further analyze its potential actual impact at the molecular interaction level. Secondly, the black tea processing control system acquires receptor protein structure models related to black tea flavor perception and matches docking ligands with the receptor protein structure models one by one, establishing corresponding docking tasks. Then, flexible docking is performed in the computational environment, allowing the docking ligands and receptor protein structure models to find more reasonable binding modes under conditions that allow for certain conformational adjustments. The binding free energy corresponding to each docking result is output. This is done to simulate the real interaction state between compound molecules and target sensory receptors as closely as possible, so that subsequent results not only reflect "whether binding occurs" but also "how strong the binding is" (for example, the binding results of different compound molecules to the same receptor protein structure model may be significantly different). Then, the black tea processing control system compares and organizes all binding free energy results, converting the binding performance of each compound molecule into a corresponding sensory intensity contribution score according to preset evaluation rules. This assigns higher weight to compounds with stronger binding interactions in the scoring results. This is done to transform the differences in molecular-level interactions into quantitative results usable for subsequent calculations, rather than treating all core differential molecular markers as having the same contribution. Finally, the black tea processing control system generates sensory weight evaluation coefficients based on the sensory intensity contribution scores of each compound molecule, and establishes a one-to-one correspondence between these coefficients and the corresponding core differential molecular markers. This allows for weighting based on the actual sensory contribution of different compound molecules during subsequent comprehensive calculations of bitterness and aroma components, ensuring that the fermentation endpoint identification results more accurately reflect the true flavor formation process of black tea.

[0051] As an example, the step of calculating the weighted conversion rate of the bitter taste component and the weighted cumulative intensity of the aroma component in the core differential molecular marker according to the sensory weight evaluation coefficient includes: weighting and summing the real-time concentrations of the bitter taste component in the core differential molecular marker during fermentation according to the sensory weight evaluation coefficient to obtain a real-time weighted bitterness value; calculating the reduction ratio of the real-time weighted bitterness value relative to the fermentation initiation stage to obtain the weighted conversion rate; weighting the real-time concentrations of the aroma component in the core differential molecular marker during fermentation according to the sensory weight evaluation coefficient to obtain a real-time weighted aroma value; and calculating the increase factor of the real-time weighted aroma value relative to the fermentation initiation stage to obtain the weighted cumulative intensity.

[0052] It should be noted that the real-time weighted bitterness value reflects the intensity of bitterness-related components at each time point during fermentation, and, combined with the weighting of sensory evaluation, accurately reflects the contribution of each component to bitterness. The initial fermentation stage refers to the initial state of black tea fermentation, usually the first point in time when fermentation begins—that is, the stage where fermentation has not yet begun or has only been underway for a very short time. The reduction ratio refers to the degree of change in the real-time weighted bitterness value relative to the initial fermentation stage. The real-time weighted aroma value reflects the intensity of changes in aroma components over time during fermentation. The increase factor refers to the degree of increase in the real-time weighted aroma value relative to the initial fermentation stage.

[0053] Understandably, firstly, the black tea processing control system multiplies the real-time concentration of bitter and astringent components in the core differential molecular markers at each fermentation time point with their corresponding sensory weights based on sensory weight evaluation coefficients. Then, it sums the weighted concentrations of all bitter and astringent components to obtain a real-time weighted bitterness value. This is done to comprehensively consider the impact of each bitter and astringent component on the overall bitterness, ensuring that the contribution of different components is reasonably reflected in the calculation. Secondly, the system calculates the reduction ratio of the real-time weighted bitterness value relative to the initial fermentation stage. By comparing the real-time weighted bitterness value at the initial fermentation stage with the real-time weighted bitterness value at the current time point, it calculates the degree of bitterness reduction (e.g., if the bitterness value at the current time point is 60% of the initial value, then the reduction ratio is 40%). This ratio reflects the degree of bitterness transformation, thus obtaining a weighted conversion rate used to assess the rate of bitterness reduction during fermentation.

[0054] Then, the black tea processing control system multiplies the real-time concentration of aroma components in the core differential molecular markers during fermentation by their sensory weights, using the same sensory weight evaluation coefficients, to obtain a real-time weighted aroma value. This is done to weight the contribution of each aroma component in conjunction with the sensory evaluation weights during aroma formation, so as to more accurately reflect the overall aroma enhancement. Finally, the system calculates the increase factor of the real-time weighted aroma value relative to the initial stage of fermentation by comparing the real-time weighted aroma value at the current time point with the real-time weighted aroma value at the initial stage of fermentation to obtain the aroma enhancement factor (for example, if the current aroma value is twice the initial value, the increase factor is 1). This factor is used to calculate the weighted cumulative intensity, reflecting the cumulative intensity of the aroma, to determine the continued enhancement of the aroma and to support the determination of the optimal fermentation endpoint.

[0055] This embodiment provides a method for identifying the fermentation endpoint of black tea based on changes in key metabolites. First, multi-source flavor characteristic data of the black tea sample under a preset fermentation time series are acquired, including sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic genetic characteristics, providing a data foundation for subsequent analysis. Next, correlation analysis is performed on these multi-source flavor characteristic data to construct a gene-metabolism-sensory multidimensional correlation network. By integrating information from different data levels, the complex mechanisms of black tea quality changes are revealed more comprehensively. Then, based on this network, metabolic components with correlation coefficients greater than a first preset correlation threshold for bitterness and astringency response values ​​and correlation coefficients greater than a second preset correlation threshold for floral and fruity aroma attribute scores are screened to obtain core differential molecular markers, accurately locating key metabolites characterizing the flavor transformation of black tea. Finally, the binding affinity of these core differential molecular markers to target sensory receptors is calculated, and sensory weight evaluation coefficients are obtained, which helps quantify the contribution of different metabolites to sensory experience and provides theoretical support for flavor optimization. Subsequently, based on sensory weight evaluation coefficients, the weighted conversion rates of bitterness components and the weighted cumulative intensities of aroma components in the core differential molecular markers were calculated. This helps monitor changes in bitterness and aroma during fermentation, ensuring an ideal balance of flavor components. Finally, when the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, the corresponding fermentation time is determined as the optimal fermentation endpoint. This embodiment can determine the optimal fermentation endpoint where bitterness and floral / fruity aromas reach a balance during black tea fermentation, ensuring the optimization and consistency of black tea flavor.

[0056] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites according to this application. Step S20 of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites includes steps S21 to S26: Step S21: Modular clustering of the transcriptome gene features is performed using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules; Step S22: Calculate the Pearson correlation coefficient between the characteristic gene value of each co-expression module and each sensory attribute in the sensory digital features, and the content of each metabolite in the metabolomic chemical features; Step S23: Filter target modules whose Pearson correlation coefficient is greater than a preset correlation threshold; Step S24: Extract hub genes with a preset connectivity from the target module and identify differential metabolites in the metabolomic chemical characteristics corresponding to the target module; Step S25: Based on the correlation calculation results between the hub gene, the differential metabolites, and the sensory attributes in the sensory digital features, establish the topological connection relationship among the three. Step S26: Gene-metabolism-sensory multidimensional association network is generated based on the topological connection relationship.

[0057] It should be noted that the weighted gene co-expression network analysis algorithm is an analytical method that constructs a weighted network based on the correlation of different gene expression changes and identifies gene sets with synergistic change patterns. It is used to identify modules and key genes related to the formation of black tea flavor from transcriptome gene characteristics. A co-expression module refers to a set of genes obtained through the weighted gene co-expression network analysis algorithm. Genes within this set show high consistency in expression trends at different fermentation time points and can therefore be considered as the same regulatory unit or the same functionally related unit. Feature gene values ​​are representative values ​​used to summarize the overall expression characteristics of a co-expression module. They characterize the overall expression level of this co-expression module in different samples or at different fermentation time points and serve as a unified characterization value for correlation analysis between this co-expression module and sensory attributes and metabolite content. Sensory attributes refer to specific evaluation items used to characterize the taste and aroma of black tea samples, including taste attributes such as bitterness, astringency, umami, and sweetness measured by an electronic tongue, and aroma attributes such as fruity, floral, sweet, woody, and grassy aromas obtained through quantitative descriptive analysis.

[0058] Metabolite content refers to the abundance, relative content, or quantitative results of various metabolites in the tested black tea sample obtained from metabolomics chemical characteristic detection. It is used to reflect the accumulation, consumption, or transformation degree of relevant chemical components at different fermentation time points. The Pearson correlation coefficient is a statistical measure used to measure the direction and strength of linear correlation between two continuous variables. The closer its value is to 1, the stronger the positive correlation; the closer it is to 0, the weaker the linear correlation; and the closer it is to -1, the stronger the negative correlation. The preset correlation threshold is a correlation judgment boundary set in advance during module screening. It is used to determine whether a certain co-expression module has a sufficiently significant correlation with sensory attributes or metabolite content. In this embodiment, it is 0.8. The target module refers to the module selected from multiple co-expression modules after correlation analysis whose Pearson correlation coefficient with sensory attributes and / or metabolite content is greater than the preset correlation threshold. This module is considered a key module closely related to the formation and quality evolution of black tea flavor.

[0059] Preset connectivity refers to the pre-defined node connection strength standard used when screening hub genes, measuring the tightness of connections between a gene and other genes within the target module. Differential metabolites refer to metabolites whose content changes significantly across different fermentation time points, reflecting those whose chemical composition has changed significantly during fermentation. Correlation calculation results are numerical analysis results obtained after calculating pairwise or multidimensional correlations between hub genes, differential metabolites, and sensory attributes, indicating the direction and strength of correlations and potential linkages among the three. Topological connectivity refers to the network connection structure established based on correlation calculation results between hub genes, differential metabolites, and sensory attributes, representing which nodes are related, the strength of their connections, and the associated paths within the overall network.

[0060] Understandably, the process begins with preprocessing the transcriptome gene characteristics. Then, the correlation between genes is calculated based on their expression trends at different fermentation time points. Modular clustering is performed accordingly, grouping genes with similar trends and strong linkages into multiple co-expression modules. This is done to consolidate the vast amount of gene expression information into several analytical units with overall change characteristics, facilitating the subsequent establishment of a correspondence with flavor changes. Secondly, characteristic gene values ​​are extracted for each co-expression module, and their correlation with each sensory attribute in the sensory digitization features and the content of each metabolite in the metabolomics chemical features is calculated to obtain the corresponding Pearson correlation coefficient. This is done to clarify which sensory changes and metabolic changes are synchronously related to each co-expression module, thereby identifying the modules truly closely related to the evolution of black tea quality.

[0061] Next, target modules with Pearson correlation coefficients greater than a preset correlation threshold were selected from the calculation results. Within each target module, genes with stronger connections were further identified as hub genes. Simultaneously, relevant differential metabolites were screened based on the corresponding metabolomic chemical characteristics of the target modules. This approach focused the analysis on the most representative gene and metabolic nodes, reducing interference from irrelevant information. Then, based on the correlation calculation results between hub genes, differential metabolites, and sensory attributes in the sensory digitization features, a topological connection relationship was established among the three. This determined which objects needed to be connected and to what extent, according to the direction and strength of their correlations. This organized the scattered correlation results into a structured relational framework. Finally, a gene-metabolism-sensory multidimensional association network was generated based on the topological connection relationship, incorporating hub genes, differential metabolites, and sensory attributes into a unified network system. This provides a direct basis for subsequent screening of core differential molecular markers and analysis of flavor formation pathways during black tea fermentation.

[0062] As an example, the step of modularly clustering the transcriptome gene features using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules includes: calculating the Pearson correlation coefficient matrix between genes based on the expression levels of each gene in the transcriptome gene features; performing exponentiation on each element in the Pearson correlation coefficient matrix according to a preset soft threshold to obtain an adjacency matrix conforming to a scale-free network distribution; calculating the topological overlap measure between genes based on the adjacency matrix to obtain a topological overlap matrix used to characterize the connectivity between genes; performing hierarchical clustering using the heterogeneity of the topological overlap matrix as a distance measure to obtain a gene clustering dendrogram; and branching the gene clustering dendrogram using a dynamic tree pruning algorithm and merging similar modules based on the correlation of characteristic genes in each branch to obtain multiple co-expression modules.

[0063] It should be noted that expression level refers to the expression level or abundance of each gene at different fermentation time points in the transcriptome gene features. It reflects the transcriptional activity of the corresponding gene in each sample and serves as the basis for subsequent calculations of gene correlations and construction of co-expression networks. The Pearson correlation coefficient matrix is ​​the result of arranging the Pearson correlation coefficients calculated pairwise based on the expression levels of each gene at multiple samples or fermentation time points in matrix form. Each element in the matrix represents the degree of consistency and direction of correlation between the expression changes of the corresponding two genes. The preset soft threshold is a power-law parameter pre-set when constructing the weighted gene co-expression network. It is used to perform a power operation on each element in the Pearson correlation coefficient matrix to strengthen the connection between strongly correlated genes and weaken the connection between weakly correlated genes, making the resulting network more consistent with the expected network structure characteristics. Strongly correlated genes refer to gene pairs in the transcriptome gene features where the expression change trends of two genes at multiple samples or fermentation time points show high consistency or strong correspondence. This is usually manifested by a large absolute value of their Pearson correlation coefficient (greater than a certain preset threshold). Weakly correlated genes refer to gene pairs where the expression trends of two genes are not very consistent across multiple samples or fermentation time points, and the correspondence is weak. This is usually manifested by a small absolute value of their Pearson correlation coefficient (less than a certain preset threshold).

[0064] Scale-free network distribution refers to a network distribution characteristic where most nodes have few connections and a few nodes have a high number of connections. It is often used to characterize the unevenness of node connections in biological networks. In this embodiment, making the network conform to a scale-free network distribution is to make the constructed gene co-expression network closer to the connection patterns of actual biological regulatory networks. The adjacency matrix is ​​a weighted connection matrix obtained by processing the Pearson correlation coefficient matrix with a preset soft threshold. Each element represents the connection strength between two corresponding genes, used to characterize whether and how strong the network connection exists between genes. Topological overlap measure is an index used to measure the degree to which two genes share adjacency relationships in the network. It considers not only whether the two genes are directly connected, but also whether they have similar connection patterns with other surrounding genes, used to more stably reflect the network similarity between genes. Intergene connectivity refers to the breadth and strength of a gene's connections with other genes in the network, used to reflect the degree of association and positional importance of the gene in the entire co-expression network. More and stronger connections usually indicate that the gene plays a more critical role in the network.

[0065] The topological overlap matrix is ​​a matrix calculated from the adjacency matrix to characterize the degree of topological overlap between any two genes. Each element represents the similarity and shared connectivity of the two genes in the network structure. Heterogeneity refers to the difference in connection patterns or topological similarity between different genes in the topological overlap matrix. In this embodiment, heterogeneity is used as a distance metric, essentially using the magnitude of the network structure differences between genes to determine whether they should be assigned to the same cluster branch. The distance metric is a calculation standard used to quantify the degree of difference between any two genes. In this step, it supports hierarchical clustering analysis, prioritizing the merging of genes with similar network structures and separating genes with significant differences. The gene clustering dendrogram is a tree-like structure diagram formed after hierarchical clustering of genes based on the distance metric, used to represent the clustering relationships between different genes from near to far. The dynamic tree pruning algorithm is a clustering method that adaptively identifies and slits branches in the gene clustering dendrogram, used to automatically identify gene branches with clear boundaries from the tree structure. Feature gene correlation refers to the degree of correlation between feature gene values ​​corresponding to different branches or different co-expression modules. It is used to determine whether different branches are sufficiently similar in overall expression patterns. When the correlation is high, the corresponding branches can be further merged into the same co-expression module.

[0066] Understandably, the process begins by organizing the expression levels of each gene in the transcriptome at different samples or fermentation time points, extracting the expression change sequences for each gene, and then calculating the correlation between any two gene expression change sequences to obtain a Pearson correlation coefficient matrix. This is done to quantify whether and to what extent a large number of genes change synchronously, providing a foundation for subsequent network construction. Secondly, each element in the Pearson correlation coefficient matrix is ​​exponentially raised according to a preset soft threshold, further strengthening highly correlated gene pairs and relatively weakening low-correlation gene pairs, thus obtaining an adjacency matrix that conforms to the scale-free network distribution. This transforms the original correlation relationships into connection relationships more suitable for network analysis, making the network structure better reflect the primary and secondary association characteristics between genes.

[0067] Then, the topological overlap measure between genes is calculated based on the adjacency matrix. This not only examines whether two genes are directly related, but also whether they are connected to the same or similar group of genes. This yields a topological overlap matrix characterizing gene connectivity. This approach reduces the randomness caused by relying on a single correlation value, making the determination of gene relationships more stable. Next, hierarchical clustering is performed using the heterogeneity of the topological overlap matrix as a distance measure. Genes with similar connection patterns are gradually merged into the same branch, while genes with more different connection patterns are gradually separated, ultimately resulting in a gene clustering dendrogram. This is done to identify which genes naturally tend to form similar variation groups from an overall structural perspective. Finally, a dynamic tree pruning algorithm is used to branch the gene clustering dendrogram. Several initial branches are identified, and branches with highly similar expression patterns are further merged based on the correlation of characteristic genes in each branch, resulting in multiple co-expression modules. This avoids overly fragmented or coarse module division, ensuring that the final co-expression modules retain internal consistency and facilitate subsequent correspondence with sensory and metabolic changes.

[0068] This embodiment first uses a weighted gene co-expression network analysis algorithm to modularly cluster transcriptome gene features, grouping genes with similar expression trends into multiple co-expression modules. This integrates scattered gene expression information into several analytical units with overall characteristics, improving the targeting of subsequent correlation analysis. Then, it calculates the Pearson correlation coefficient between the characteristic gene values ​​of each co-expression module and the sensory attributes in the sensory digital features, as well as the content of each metabolite in the metabolomic chemical features. This clarifies the correspondence between different co-expression modules and flavor expression and metabolic changes, facilitating the identification of modules more closely related to the evolution of black tea quality. Next, it screens target modules with Pearson correlation coefficients greater than a preset correlation threshold, retaining those with strong correlations to reduce interference from irrelevant information. Subsequently... By extracting hub genes with preset connectivity from the target module and identifying differential metabolites in the metabolomic chemical features corresponding to the target module, representative key genes and key metabolic nodes in the network can be obtained, which helps to highlight the core changing factors in the flavor formation process. Furthermore, based on the correlation calculation results between hub genes, differential metabolites, and sensory attributes in the sensory digital features, a topological connection relationship among the three is established, thereby organizing the originally scattered correlation results into a structured connection framework, which is conducive to clearly characterizing the linkage relationship among the three. Finally, a gene-metabolism-sensory multidimensional association network is generated based on the topological connection relationship, revealing the synergistic relationship between gene changes, metabolic changes, and sensory changes in the fermentation process of black tea as a whole, providing a basis for subsequent screening of core differential molecular markers.

[0069] The transcriptome gene features are modularly clustered using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules. The Pearson correlation coefficients between the feature gene values ​​of each co-expression module and the sensory attributes in the sensory digital features, and the content of each metabolite in the metabolomic chemical features, are calculated. Target modules with Pearson correlation coefficients greater than a preset correlation threshold are selected. Hub genes with preset connectivity are extracted from the target modules, and differentially expressed metabolites in the metabolomic chemical features corresponding to the target modules are identified. Based on the correlation calculation results between the hub genes, the differentially expressed metabolites, and the sensory attributes in the sensory digital features, a topological connection relationship is established among the three. A gene-metabolism-sensory multidimensional association network is generated based on the topological connection relationship.

[0070] This application also provides a device for identifying the fermentation endpoint of black tea based on changes in key metabolites. Please refer to [link / reference]. Figure 3 The black tea fermentation endpoint identification device based on key metabolite changes includes: Data acquisition module 10 is used to acquire multi-source flavor feature data of the black tea sample to be tested under a preset fermentation time series. The multi-source flavor feature data includes sensory digital features, metabolomic chemical features and transcriptomic gene features. The association network construction module 20 is used to perform association analysis on the multi-source flavor feature data and construct a gene-metabolism-sensory multidimensional association network characterizing the evolution of black tea quality. The marker screening module 30 is used to screen metabolic components with a correlation coefficient greater than a first preset correlation threshold and a correlation coefficient greater than a second preset correlation threshold with the flower and fruit aroma attribute score based on the gene-metabolism-sensory multidimensional association network, so as to obtain core differential molecular markers. The affinity calculation module 40 is used to calculate the binding affinity between the core differential molecular marker and the target sensory receptor to obtain the sensory weight evaluation coefficient. The weight quantization module 50 is used to calculate the weighted conversion rate of the bitter taste component and the weighted cumulative intensity of the aroma component in the core differential molecular markers based on the sensory weight evaluation coefficients. The endpoint determination module 60 is used to determine the optimal fermentation endpoint when the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value.

[0071] The black tea fermentation endpoint identification device based on key metabolite changes provided in this application, employing the black tea fermentation endpoint identification method based on key metabolite changes in the above embodiments, can solve the technical problem of how to determine the optimal fermentation endpoint for achieving a balance between bitterness and floral / fruity aroma during black tea fermentation. Compared with the prior art, the beneficial effects of the black tea fermentation endpoint identification device based on key metabolite changes provided in this application are the same as those of the black tea fermentation endpoint identification method based on key metabolite changes provided in the above embodiments, and other technical features in the black tea fermentation endpoint identification device based on key metabolite changes are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0072] This application provides a device for identifying the fermentation endpoint of black tea based on changes in key metabolites. The device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method for identifying the fermentation endpoint of black tea based on changes in key metabolites in the first embodiment described above.

[0073] The following is for reference. Figure 4The diagram illustrates a structural schematic of a black tea fermentation endpoint identification device based on key metabolite changes, suitable for implementing embodiments of this application. The black tea fermentation endpoint identification device based on key metabolite changes in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Android Devices), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The illustrated device for identifying the fermentation endpoint of black tea based on changes in key metabolites is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0074] like Figure 4 As shown, the black tea fermentation endpoint identification device based on key metabolite changes may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in ROM 1002 (Read Only Memory) or a program loaded from storage device 1003 into RAM 1004 (Random Access Memory). RAM 1004 also stores various programs and data required for the operation of the oil depot fire simulation scenario construction device that integrates dual-model inversion. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the black tea fermentation endpoint identification device based on key metabolite changes to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a black tea fermentation endpoint identification device based on key metabolite changes with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0075] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0076] The black tea fermentation endpoint identification device based on key metabolite changes provided in this application, employing the black tea fermentation endpoint identification method based on key metabolite changes in the above embodiments, can solve the technical problem of how to determine the optimal fermentation endpoint for achieving a balance between bitterness and floral / fruity aroma during black tea fermentation. Compared with the prior art, the beneficial effects of the black tea fermentation endpoint identification device based on key metabolite changes provided in this application are the same as those of the black tea fermentation endpoint identification method based on key metabolite changes provided in the above embodiments, and other technical features in this black tea fermentation endpoint identification device based on key metabolite changes are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0077] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0078] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0079] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the method for identifying the fermentation endpoint of black tea based on changes in key metabolites in the above embodiments.

[0080] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, RAM (Random Access Memory), ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory or Flash Memory), optical fibers, CD-ROM (CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0081] The aforementioned computer-readable storage medium may be included in a black tea fermentation endpoint identification device based on key metabolite changes; or it may exist independently and not assembled into a black tea fermentation endpoint identification device based on key metabolite changes.

[0082] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the black tea fermentation endpoint identification device based on key metabolite changes, the black tea fermentation endpoint identification device based on key metabolite changes causes the device to: acquire multi-source flavor characteristic data of the black tea sample under test at a preset fermentation time series, wherein the multi-source flavor characteristic data includes sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic gene characteristics; perform correlation analysis on the multi-source flavor characteristic data to construct a gene-metabolism-sensory multidimensional correlation network characterizing the evolution of black tea quality; and, based on the gene-metabolism-sensory multidimensional correlation network, screen for flavors related to bitterness and astringency. Metabolic components whose correlation coefficient with the taste response value is greater than a first preset correlation threshold and whose correlation coefficient with the floral and fruity aroma attribute score is greater than a second preset correlation threshold are identified as core differential molecular markers. The binding affinity of the core differential molecular markers to the target sensory receptors is calculated to obtain sensory weight evaluation coefficients. Based on the sensory weight evaluation coefficients, the weighted conversion rate of bitter taste components and the weighted cumulative intensity of aroma components in the core differential molecular markers are calculated respectively. When the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, the corresponding fermentation time is taken as the optimal fermentation endpoint.

[0083] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including LAN (Local Area Network) or WAN (Wide Area Network)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0084] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0085] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described method for identifying the fermentation endpoint of black tea based on changes in key metabolites. This method solves the technical problem of determining the optimal fermentation endpoint where bitterness and floral / fruity aromas reach a balance during black tea fermentation. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the black tea fermentation endpoint identification method based on changes in key metabolites provided in the above embodiments, and will not be elaborated upon here.

[0086] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites as described above.

[0087] The computer program product provided in this application solves the technical problem of determining the optimal fermentation endpoint for achieving a balance between bitterness and floral / fruity aroma during the fermentation process of black tea. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the black tea fermentation endpoint identification method based on key metabolite changes provided in the above embodiments, and will not be repeated here.

[0088] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for identifying the fermentation endpoint of black tea based on changes in key metabolites, characterized in that, The method includes: Acquire multi-source flavor characteristic data of the black tea sample to be tested under a preset fermentation time series. The multi-source flavor characteristic data includes sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic gene characteristics. Correlation analysis was performed on the multi-source flavor characteristic data to construct a gene-metabolism-sensory multidimensional correlation network characterizing the evolution of black tea quality; Based on the gene-metabolism-sensory multidimensional association network, metabolic components with a correlation coefficient greater than the first preset correlation threshold and a correlation coefficient greater than the second preset correlation threshold with the floral and fruity aroma attribute score were screened to obtain core differential molecular markers. The binding affinity of the core differential molecular markers to the target sensory receptors is calculated to obtain the sensory weight evaluation coefficient; Based on the sensory weight evaluation coefficients, the weighted conversion rate of bitter components and the weighted cumulative intensity of aroma components in the core differential molecular markers were calculated respectively. When the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value, the corresponding fermentation time is taken as the optimal fermentation endpoint.

2. The method as described in claim 1, characterized in that, The steps of performing correlation analysis on the multi-source flavor characteristic data to construct a gene-metabolism-sensory multidimensional correlation network characterizing the evolution of black tea quality include: The transcriptome gene features were modularly clustered using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules. Calculate the Pearson correlation coefficient between the characteristic gene value of each co-expression module and each sensory attribute in the sensory digital features, and the content of each metabolite in the metabolomic chemical features; Target modules whose Pearson correlation coefficient is greater than a preset correlation threshold are selected; Hub genes with a preset connectivity are extracted from the target module, and differential metabolites in the metabolomic chemical characteristics corresponding to the target module are identified. Based on the correlation calculation results among the hub gene, the differential metabolites, and the sensory attributes in the sensory digitization features, a topological connection relationship among the three is established. Gene-metabolism-sensory multidimensional network is generated based on the topological connections.

3. The method as described in claim 2, characterized in that, The step of modularly clustering the transcriptome gene features using a weighted gene co-expression network analysis algorithm to obtain multiple co-expression modules includes: Based on the expression levels of each gene in the transcriptome gene characteristics, calculate the Pearson correlation coefficient matrix between genes; The Pearson correlation coefficient matrix is ​​exponentially operated on each element according to a preset soft threshold to obtain an adjacency matrix that conforms to the scale-free network distribution. The topological overlap measure between genes is calculated based on the adjacency matrix to obtain the topological overlap matrix used to characterize the connectivity between genes; Using the heterogeneity of the topological overlap matrix as a distance metric, hierarchical clustering is performed to obtain a gene clustering dendrogram; The gene clustering dendrogram is branched using a dynamic tree pruning algorithm, and similar modules are merged based on the correlation of characteristic genes in each branch to obtain multiple co-expression modules.

4. The method as described in claim 1, characterized in that, The step of obtaining multi-source flavor characteristic data of the black tea sample under a preset fermentation time series, wherein the multi-source flavor characteristic data includes sensory digital characteristics, metabolomic chemical characteristics, and transcriptomic genetic characteristics, includes: The taste response potential values ​​of the black tea sample to be tested are collected by a taste sensor to obtain a taste feature vector containing bitter, astringent, fresh and sweet response values. The response signal intensity of the black tea sample under test in the headspace environment is collected by a gas-sensitive sensor array, and the response signal intensity is matrixed according to the sensor channel dimension to obtain the odor feature vector. The taste feature vector and the odor feature vector are concatenated and fused to obtain sensory digital features; The non-volatile and volatile compounds in the black tea sample were qualitatively and quantitatively analyzed by liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry, respectively, to obtain the metabolomic chemical characteristics. Transcript abundance information of the black tea sample under test at various time points was obtained by high-throughput sequencing to obtain transcriptome gene characteristics. The sensory digital features, metabolomic chemical features, and transcriptomic gene features are aligned along a timeline according to a preset fermentation time sequence to obtain multi-source flavor feature data.

5. The method as described in claim 1, characterized in that, The step of screening metabolic components with correlation coefficients greater than a first preset correlation threshold and correlation coefficients greater than a second preset correlation threshold with floral and fruity aroma attribute scores based on the gene-metabolism-sensory multidimensional association network to obtain core differential molecular markers includes: Bitter taste components with a correlation coefficient greater than a first preset correlation threshold are extracted from the gene-metabolism-sensory multidimensional association network. The bitter taste components include epigallocatechin gallate. Aroma components with correlation coefficients greater than a second preset correlation threshold with floral and fruity aroma attribute scores are extracted from the gene-metabolism-sensory multidimensional association network. The aroma components include geraniol, damastone, and phenylacetaldehyde. Amino acid components that are positively correlated with the umami and sweetness response value were extracted from the gene-metabolism-sensory multidimensional association network, and the amino acid components included alanyl-glutamine. From the gene-metabolism-sensory multidimensional association network, pivot genes that have co-expression relationships with the bitter taste component, the aroma component and the amino acid component are screened, and the pivot genes include deubiquitinase genes; The core differential molecular markers were obtained by summarizing the bitter taste components, the aroma components, the amino acid components, and the hub gene.

6. The method as described in claim 1, characterized in that, The step of calculating the binding affinity between the core differential molecular marker and the target sensory receptor to obtain the sensory weight evaluation coefficient includes: The three-dimensional chemical structures of each compound molecule in the core differential molecular markers were obtained as docking ligands; Obtain structural models of receptor proteins, including bitter taste receptors, sweet taste receptors, and olfactory receptors. The docking ligand is flexibly docked with the receptor protein structure model, and the binding free energy is calculated. The sensory intensity contribution of each compound molecule is scored based on the magnitude of its binding free energy, resulting in a sensory weight evaluation coefficient.

7. The method according to any one of claims 1 to 6, characterized in that, The step of calculating the weighted conversion rate of the bitter taste component and the weighted cumulative intensity of the aroma component in the core differential molecular markers based on the sensory weight evaluation coefficients includes: The real-time weighted bitterness value is obtained by weighting and summing the real-time concentrations of bitter components in the core differential molecular markers during fermentation based on the sensory weight evaluation coefficients. The weighted conversion rate is obtained by calculating the reduction ratio of the real-time weighted bitterness value relative to the initial stage of fermentation. The real-time concentrations of aroma components in the core differential molecular markers during fermentation are weighted according to the sensory weight evaluation coefficients to obtain real-time weighted aroma values. The weighted cumulative intensity is obtained by calculating the increase factor of the real-time weighted aroma value relative to the fermentation initiation stage.

8. A device for identifying the fermentation endpoint of black tea based on changes in key metabolites, characterized in that, The device includes: The data acquisition module is used to acquire multi-source flavor feature data of the black tea sample to be tested under a preset fermentation time series. The multi-source flavor feature data includes sensory digital features, metabolomic chemical features and transcriptomic gene features. The association network construction module is used to perform association analysis on the multi-source flavor feature data and construct a gene-metabolism-sensory multidimensional association network characterizing the evolution of black tea quality. The marker screening module is used to screen metabolic components with a correlation coefficient greater than a first preset correlation threshold and a correlation coefficient greater than a second preset correlation threshold with the floral and fruity aroma attribute score based on the gene-metabolism-sensory multidimensional association network, thereby obtaining core differential molecular markers. The affinity calculation module is used to calculate the binding affinity between the core differential molecular marker and the target sensory receptor to obtain the sensory weight evaluation coefficient. The weight quantization module is used to calculate the weighted conversion rate of bitter taste components and the weighted cumulative intensity of aroma components in the core differential molecular markers based on the sensory weight evaluation coefficients. The endpoint determination module is used to determine the optimal fermentation endpoint when the rate of change of the weighted conversion rate is less than a preset rate of change threshold and the weighted cumulative intensity is greater than or equal to a preset maximum peak value.

9. A device for identifying the fermentation endpoint of black tea based on changes in key metabolites, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the method for identifying the fermentation endpoint of black tea based on changes in key metabolites as described in any one of claims 1 to 7.

Images