A method for evaluating the post-slaughter storage time of mutton based on multi-omics data

Abstract

This application discloses a method for assessing the post-slaughter storage time of mutton based on multi-omics data, comprising: collecting longissimus dorsi muscle samples from Kazakh sheep at multiple known time points after slaughter; acquiring quantitative data through proteomics, metabolomics, and volatile metabolomics technologies to construct a multi-omics time series matrix; employing a two-level comparison and exclusion screening strategy to screen time-specific core biomarkers; determining individualized time decay rates using a first-level kinetic decay model; obtaining support scores for each omics channel based on a dynamic confidence weight model to determine preliminary judgment results; calculating a cross-omics conflict index; and initiating an arbitration mechanism to adjust channel weights when the conflict index is below a threshold; and determining the assessment result based on channel weights and support scores. This application achieves accurate assessment of the post-slaughter storage time of mutton by dynamically quantifying the time decay characteristics of biomarkers, intelligently resolving multi-omics evidence conflicts, and outputting a confidence rating.

Description

Technical Field

[0001] This application relates to the field of food quality and safety testing technology, and relates to, but is not limited to, a method for evaluating the post-slaughter storage time of mutton based on multi-omics data. Background Technology

[0002] Accurate assessment of post-slaughter storage time for livestock meat is a core aspect of meat quality control and traceability management. Traditional methods mainly rely on sensory evaluation or single physicochemical index testing, which has drawbacks such as strong subjectivity, low temporal resolution (usually only able to distinguish between "fresh" and "not fresh"), and inability to adapt to breed specificity.

[0003] With the development of omics technologies, multi-omics data fusion has provided new insights for post-mortem time assessment. However, existing methods often employ static threshold judgments or static models for time assessment, neglecting the information entropy decay characteristics of biomarkers over storage time. Under long-term storage conditions, static models experience a significant decrease in temporal resolution as biomarker concentrations tend to stabilize or degrade to baseline levels, leading to a substantial reduction in the assessment accuracy of long-stored samples.

[0004] Furthermore, discrepancies often arise between multi-source data such as proteomics, metabolomics, and volatile metabolomics, and existing technologies mostly employ simple weighted fusion methods to process multi-omics data. This fusion strategy is insufficient to effectively resolve contradictory evidence, and the varying reliability of different omics data after long-term storage can lead to error accumulation, further reducing the accuracy and stability of the assessment results. Simultaneously, existing methods lack quantitative credibility indicators, failing to provide support for hierarchical decision-making.

[0005] For local breeds such as Kazakh sheep, existing technologies do not take into account breed-specific metabolic kinetic differences. Directly applying the evaluation model for ordinary sheep will result in significant deviations, making it difficult to meet the industry's demand for accurate traceability and quality grading of high-end mutton products. Summary of the Invention

[0006] To address the problems in existing technologies, such as the significant decrease in the accuracy of static models under long-term storage conditions, the accumulation of errors due to simple weighted fusion of multi-source data, the lack of quantitative credibility indicators to support hierarchical decision-making, and the failure to consider the differences in metabolic kinetics specific to different breeds, this application provides a method for evaluating the post-slaughter storage time of mutton based on multi-omics data.

[0007] The technical solution of this application embodiment is implemented as follows: This application embodiment provides a method for evaluating the post-slaughter storage time of mutton based on multi-omics data, the method comprising: Samples of the longissimus dorsi muscle were collected from Kazakh sheep at multiple known time points after slaughter. Quantitative data of proteins, metabolites, and volatile compounds were obtained using proteomics, metabolomics, and volatile metabolomics technologies, respectively. After standardization, a multi-omics time series matrix was constructed. Some of the known time points were used as a candidate time set. For the multi-omics time series matrix, a pre-constructed nested strategy of two-level comparison and exclusion screening is adopted to screen out multiple time-specific core biomarkers; the individualized time decay rate of each time-specific core biomarker is determined by a pre-constructed first-level kinetic decay model. Based on the individualized time decay rate and the pre-built dynamic confidence weight model, the support score of each omics channel to the candidate time set is obtained; based on the support score, the preliminary judgment result of each omics channel is determined; based on the preliminary judgment result of each omics channel, the cross-omics conflict index is calculated; when the cross-omics conflict index is lower than a preset threshold, the conflict arbitration mechanism is activated to determine the weight of each omics channel. The evaluation results of the post-slaughter storage time of mutton are determined based on the weights of each omics channel and the support scores.

[0008] In some embodiments, for the multi-omics time series matrix, a pre-constructed nested strategy of two-level comparison and exclusion screening is used to screen out multiple time-specific core biomarkers, including: The first specific difference between 24-hour and 1-hour quantitative data in the multi-omics time series matrix is ​​determined, and quantitative data with a first specific difference greater than a preset first specific difference threshold are selected as the first specific core biomarker set. The second specific difference between 72-hour and 1-hour quantitative data in the multi-omics time series matrix, and the third specific difference between 72-hour and 24-hour quantitative data in the multi-omics time series matrix are determined. Quantitative data satisfying both the second specific difference and the third specific difference being greater than a preset second specific difference threshold are selected as the second specific core biomarker set. Common biomarkers between the first and second specific core biomarker sets are removed, and the remaining biomarkers are merged to obtain multiple time-specific core biomarkers. The multiple time-specific core biomarkers include: proteomics biomarkers, metabolomics biomarkers, and volatile substance metabolomics biomarkers.

[0009] In some embodiments, the pre-construction process of the first-order dynamic decay model includes: Based on the multiple time-specific core biomarkers, the relative concentration sequence of each time-specific core biomarker was determined; For each time-specific core biomarker, starting from the initial concentration at the obtained baseline sampling time point, and using the standardized concentration at the obtained known time point as the dependent variable, and with the known time point as the independent variable, a nonlinear least squares method is used to fit the exponential decay curve to obtain a first-order kinetic decay model.

[0010] In some embodiments, the first-order dynamic decay model satisfies the following formula: in, For the first A time-specific core biomarker at a known time point Standardized concentration, For the first Several time-specific core biomarkers at the baseline sampling time point The initial concentration, For the first Background baseline values ​​for the concentration detection of time-specific core biomarkers. It is a natural constant. For the first The time decay rate of a time-specific core biomarker.

[0011] In some embodiments, the omics channels include proteomics channels, metabolomics channels, and volatile substance metabolomics channels; The support score for each omics channel to the candidate time set is obtained based on the individualized time decay rate and the pre-constructed dynamic confidence weight model, including: The individualized time decay rate is input into a pre-built dynamic confidence weight model to obtain the dynamic confidence weight. Based on the dynamic confidence weights, the support scores for each candidate time point are determined for the proteomics channel, the metabolomics channel, and the volatile metabolomics channel; based on the support scores for each candidate time point, the support scores for each omics channel to the candidate time set are determined.

[0012] In some embodiments, the pre-construction process of the dynamic confidence weight model includes: Based on the individualized time decay rate, the basic weights of the omics channels are determined; Based on the weights of each unknown storage time point to be tested, species correction factors and temperature correction factors are determined; Based on the basic weights of the omics channels, the species correction factor, and the temperature correction factor, a dynamic confidence weight model is constructed. The dynamic confidence weight model satisfies the following formula: in, For the first A time-specific core biomarker at an unknown storage time. Dynamic confidence weights, The storage time to be tested is unknown. The baseline weights for each time-specific core biomarker; The time delay is the time difference between the unknown storage time to be tested and the baseline sampling time. For the first The time decay rate of a time-specific core biomarker As the basic weights for omics channels, As a species correction factor, This is a temperature correction factor; When the unknown storage time to be tested Equal to candidate time At that time, the dynamic confidence weight calculated by the dynamic confidence weight model is the first... A time-specific core biomarker at candidate time Confidence weight ( ).

[0013] In some embodiments, determining the post-slaughter storage time assessment result of mutton based on the weights of each omics channel and the support score includes: Based on the weights of each omics channel and the support scores, the overall confidence score for each candidate time is calculated. After probability normalization of the comprehensive confidence score, a three-level confidence rating is performed by combining the cross-omics conflict index and the probability value to determine the evaluation result of the post-slaughter storage time of mutton with quantitative confidence. The comprehensive confidence score satisfies the following formula: in, To calculate the overall confidence score, For omics channel, As a time-specific core biomarker, It is the set of time-specific core biomarkers contained in the g-th omics channel. For the first A time-specific core biomarker at candidate time Confidence weights; For the first A time-specific core biomarker at candidate time The indicator function value, when the first A time-specific core biomarker at candidate time When detected, the indicator function value is 1; otherwise, it is 0. For the g-th omics channel weight, After screening ( The total number of time-specific core biomarkers with a value ≥ 0.3.

[0014] In some embodiments, when the cross-omics conflict index is lower than a preset threshold, initiating a conflict arbitration mechanism to determine the weights of each omics channel includes: When the cross-omics conflict index is lower than a preset threshold, the standard deviation of the determination results of each time-specific core biomarker in each omics channel is determined. Based on the standard deviation, the dispersion of time-specific core biomarkers within each omics channel is determined; Based on the degree of dispersion of time-specific core biomarkers within each omics channel, the internal consistency coefficient of each omics channel is determined. Based on the internal consistency coefficient, the basic weights of the omics channels are adjusted to obtain the adjusted weights of each omics channel.

[0015] In some embodiments, the step of calculating the cross-omics conflict index based on the preliminary determination results of each omics channel includes: converting the preliminary determination results of the proteomics channel, metabolomics channel and volatile substance metabolomics channel in the omics channel into hours, respectively, to obtain the preliminary determination results of the hours of each omics channel; Based on the preliminary determination results of the number of hours of the proteomics channel, the metabolomics channel, and the volatile substance metabolomics channel, the cross-omics conflict index is calculated according to the pre-constructed cross-omics conflict index model. The cross-omics conflict index model satisfies the following formula: Wherein, β is the cross-omics conflict index. This is a preliminary result for determining the number of hours in the proteomics channel. This is a preliminary result for determining the number of hours in the metabolomics pathway. Preliminary results for determining the number of hours in the metabolomics pathways of volatile substances. The maximum time span is preset.

[0016] In some embodiments, after probabilistically normalizing the comprehensive confidence score, a three-level confidence rating is performed by combining the cross-omics conflict index and the probability value to determine the assessment result of the post-slaughter storage time of mutton with quantitative confidence, including: The comprehensive confidence scores of each candidate time point are normalized to obtain the normalized probability value of each candidate time point; the normalized probability value is between 0 and 1, and the sum of the normalized probability values ​​of all candidate time points is 1. The normalized probability values ​​of each candidate time point are sorted from largest to smallest and denoted as the first probability value, the second probability value, and so on. Based on the normalized probability value, the evaluation result is output according to the following three-level judgment rules: The first-level determination is a clear determination: when the first probability value is greater than the first probability threshold, and the difference between the first probability value and the second probability value is greater than the preset difference threshold, it is determined as the candidate time point corresponding to the first probability value, and the post-slaughter storage time of the determined time point is output. The second-level determination is an interval determination: when the first probability value is greater than the second probability threshold, and the difference between the first probability value and the second probability value is less than or equal to the preset difference threshold, it is determined that the candidate time point corresponding to the first probability value and the candidate time point corresponding to the second probability value form an adjacent time interval, and the post-slaughter storage time within this interval is output. Level 3 determination: When the first probability value is less than the third probability threshold, it is determined that no valid conclusion can be drawn, and the result of "cannot be determined" is output and a prompt for resampling and detection is given.

[0017] The beneficial effects of the technical solutions provided in this application include at least the following: This application overcomes the problems of low time resolution and strong subjectivity of single physicochemical indicators by collecting longissimus dorsi muscle samples from Kazakh sheep at multiple known time points after slaughter and constructing a multi-omics time series matrix by integrating proteomics, metabolomics and volatile metabolomics data.

[0018] This application employs a nested strategy of two-level comparison and exclusive screening to screen time-specific core biomarkers, and uses a first-level kinetic decay model to determine individualized time decay rates, thereby achieving dynamic quantification of the information entropy decay characteristics of biomarkers over time. This solves the technical problem that the accuracy of static models decreases significantly under long-term storage conditions.

[0019] This application effectively resolves the judgment conflict between multi-source data by calculating the cross-omics conflict index and initiating a conflict arbitration mechanism to adjust the weight of each omics channel, avoids the error accumulation problem caused by simple weighted fusion, and improves the accuracy and stability of the evaluation results.

[0020] This application determines the post-slaughter storage time assessment results based on the weights and support scores of each omics channel, and can output a graded assessment conclusion with quantitative credibility, providing a reliable decision-making basis for the accurate traceability and quality grading of high-end mutton products.

[0021] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the technical solutions provided in the embodiments of the present invention. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 A flowchart illustrating a method for evaluating the post-slaughter storage time of mutton based on multi-omics data, provided in an embodiment of this application; Figure 2 A schematic diagram illustrating the calculation process of the support score for a method for evaluating the post-slaughter storage time of mutton based on multi-omics data, provided in an embodiment of this application. Figure 3 A flowchart illustrating the conflict arbitration mechanism of a method for assessing post-slaughter storage time of mutton based on multi-omics data, provided in this application embodiment. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. The following embodiments are used to illustrate this application, but are not intended to limit the scope of this application. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0025] It should be noted that the terms "first, second, and third" used in the embodiments of this application are merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first, second, and third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0026] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments of this application pertain. It should also be understood that terms such as those defined in general dictionaries should be understood to have a meaning consistent with their meaning in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0027] See Figure 1 The diagram shown is a flowchart illustrating a method for evaluating the post-slaughter storage time of mutton based on multi-omics data, as provided in this application embodiment. Figure 1 The method is described below: The method may include the following steps: Step 1: Collect longissimus dorsi muscle samples from Kazakh sheep at multiple known time points after slaughter. Quantitative data of proteins, metabolites, and volatile compounds are obtained using proteomics, metabolomics, and volatile matter metabolomics technologies, respectively. After standardization, a multi-omics time series matrix is ​​constructed. Some of the known time points are used as a candidate time set. Step 2: For the multi-omics time series matrix, a pre-constructed nested strategy of two-level comparison and exclusion screening is used to screen out multiple time-specific core biomarkers; the individualized time decay rate of each time-specific core biomarker is determined by a pre-constructed first-level kinetic decay model. Step 3: Based on the individualized time decay rate and the pre-built dynamic confidence weight model, obtain the support score of each omics channel for the candidate time set; determine the preliminary judgment result of each omics channel according to the support score; calculate the cross-omics conflict index according to the preliminary judgment result of each omics channel; when the cross-omics conflict index is lower than the preset threshold, start the conflict arbitration mechanism to determine the weight of each omics channel. Step 4: Determine the evaluation result of the post-slaughter storage time of mutton based on the weights of each omics channel and the support score.

[0028] In one specific embodiment of this example, in step 1 above, healthy Kazakh sheep can be selected, and after standardized slaughter, the longissimus dorsi muscle can be taken as the sample to be tested. (1h) (12h) (24h) (48h) (72h) Samples were taken at five known time points, with n≥6 biological replicates at each known time point, and the sample mass was 50±5g. After sampling, the samples were immediately flash-frozen in liquid nitrogen and transferred to an ultra-low temperature freezer at -80℃ to avoid RNA (ribonucleic acid) degradation and protein denaturation. Frozen samples were ground in liquid nitrogen and passed through a 40-mesh sieve → pre-cooled PBS (Phosphate-Buffered Saline, pH 7.4, containing protease inhibitors) was added at a 1:4 mass-to-volume ratio → centrifuged at 12000g for 15min at 4℃ → the middle clear liquid was collected and aliquoted for proteomics and metabolomics detection, respectively. For volatile metabolomics, samples from the longissimus dorsi muscle were directly processed.

[0029] Quantitative data on proteins, metabolites, and volatile compounds were obtained using proteomics, metabolomics, and volatile matter metabolomics technologies, as detailed below: 1. Proteomics: Data-Independent Acquisition (DIA) mode was used. After extraction with SDT lysis buffer, quantification was performed using the BCA (Bicinchoninic Acid Assay) method (adjusted to 1 μg / μL), followed by enzymatic digestion using the FASP (Filter-Aided Sample Preparation) method. Separation was achieved by AcclaimPepMap RSLC C18 column liquid chromatography and detection by Orbitrap Fusion Lumos mass spectrometry. Based on the sheep proteome database (Uniprot) and Spectronaut Pulsar software, the FDR (False Discovery Rate) was <1%, and relative quantitative data of ≥2000 proteins were finally obtained. 2. Metabolomics: Gas chromatography-time of flight mass spectrometry (GC-TOF-MS) was used. After extraction with methanol:water (3:1) extractant, oxime and silanization derivatization were performed. The metabolites were then detected by DB-5MS capillary column gas chromatography-time of flight mass spectrometry. Based on the NIST2017 (National Institute of Standards and Technology) spectral library (positive and negative match >800), peak area quantitative data of ≥800 metabolites were finally obtained.

[0030] 3. Volatile Metabolomics: Solid Phase Microextraction-Gas Chromatography-Mass Spectrometry (SPME-GC-MS) was used. Muscle samples were granulated to 2 mm size and treated with saturated sodium chloride solution. Then, solid phase microextraction was performed using a 50 / 30 μm DVB / CAR / PDMS (Divinylbenzene / Carboxen / Polydimethylsiloxane) three-phase extraction head. Detection was performed using HP-INNOWax polar chromatography column gas chromatography + mass spectrometry. Identification was based on NIST 2017 + Wiley dual library (match > 800), and semi-quantitative analysis was performed using internal standard method. Finally, relative quantitative data of ≥60 volatile compounds were obtained.

[0031] The quantitative data from the three omics studies mentioned above are then integrated along the time dimension, defining the sub-matrix and the integration tensor. Proteomics matrix. (p≥2000, 5 indicates known sampling time points) - Quantity); Metabolome matrix (m≥800); volatile metabolome matrix (f≥60); integrated into a multi-omics time series matrix .

[0032] For any element of a multi-omics time series matrix The detection value of the i-th biomarker at time j is normalized using Z-score to eliminate dimensional differences between omics, ensuring that the data follows a standard normal distribution with a mean of 0 and a standard deviation of 1. If a biomarker is missing in >30% of the samples, it is removed; the remaining missing values ​​are filled using K-nearest neighbor imputation (k=5), finally yielding a standardized, missing-free multi-omics time series matrix.

[0033] Furthermore, step 2 above, which describes using a pre-constructed nested strategy of two-level comparison and exclusion screening to select multiple time-specific core biomarkers for the multi-omics time series matrix, may include the following steps: Step 2.1: Determine the first specific difference between the 24-hour quantitative data and the 1-hour quantitative data in the multi-omics time series matrix, and select the quantitative data with the first specific difference greater than a preset first specific difference threshold as the first specific core biomarker set; determine the second specific difference between the 72-hour quantitative data and the 1-hour quantitative data in the multi-omics time series matrix, and the third specific difference between the 72-hour quantitative data and the 24-hour quantitative data in the multi-omics time series matrix; select the quantitative data that satisfy the second specific difference being greater than a preset second specific difference threshold and the third specific difference being greater than a preset third specific difference threshold as the second specific core biomarker set; Step 2.2: Remove the common biomarkers in the first specific core biomarker set and the second specific core biomarker set, and merge the remaining biomarkers to obtain multiple time-specific core biomarkers; The multiple time-specific core biomarkers include: proteomics biomarkers, metabolomics biomarkers, and volatile substance metabolomics biomarkers.

[0034] In one specific embodiment of this example, in step 2.1 above, the first specific difference, the second specific difference, and the third specific difference are all quantified using the difference fold as an indicator, and the significance level is calculated in conjunction with statistical tests. The specific screening process includes the following steps: Step 2.1.1: For each quantitative data point (including protein, metabolite or flavor compound) in the multi-omics time series matrix, calculate the fold difference between the two time points of 24 hours and 1 hour post-mortem, and assess the statistical significance of the difference; The first specificity difference threshold can be set as a difference fold of 2. Features that meet the first specificity difference threshold are included in the first specificity core biomarker set. This set of first-specific core biomarkers characterizes omics features that change significantly at 24 hours post-mortem compared to 1 hour, primarily reflecting biochemical processes such as early post-mortem muscle rigidity and glycolysis initiation.

[0035] Step 2.1.2: Calculate the fold difference of each quantitative data point between 72 hours and 1 hour post-slaughter, and between 72 hours and 24 hours post-slaughter, and assess their statistical significance.

[0036] The second specificity threshold can be set as follows: the fold change between 72 hours and 1 hour post-mortem reaches approximately 2.83 times, and the fold change between 72 hours and 24 hours post-mortem reaches 2 times. If both fold change conditions are met, and the statistical tests for both groups are statistically significant (p-value less than 0.05), the marker is included in the set of core second specificity biomarkers. .

[0037] Considering the technical characteristics and biological variability of different omics data, this embodiment employs differentiated screening thresholds for the proteome, metabolome, and volatile metabolome: for the volatile metabolome, the two fold change thresholds are relaxed to approximately 2.3-fold and approximately 1.74-fold, respectively. This second set of specific core biomarkers characterizes omics features that undergo sustained and significant changes at 72 hours post-mortem compared to earlier times (1 hour and 24 hours), primarily reflecting biochemical processes such as protein degradation, ATP depletion, and lipid oxidation in the later post-mortem period.

[0038] The first set of specific core biomarkers obtained in step 2.1.1 above is compared with the second set of specific core biomarkers obtained in step 2.1.2. Biomarkers shared by both sets (features that change significantly in both 24 and 72 hours) are removed, and biomarkers that are specific only within a single time window (24 or 72 hours) are retained. The final result is a time-specific core biomarker set, which can be expressed as follows: , ,final and This could be multiple time-specific core biomarkers; for example, there could be 12 time-specific core biomarkers (including 7 proteins, 3 metabolites, and 2 flavor compounds).

[0039] The pre-construction process of the first-order dynamic decay model mentioned in step 2 above may include the following steps: Step a: Based on the multiple time-specific core biomarkers, determine the relative concentration sequence of each time-specific core biomarker; Step b: For each time-specific core biomarker, starting from the initial concentration at the obtained baseline sampling time point, and using the standardized concentration at the obtained known time point as the dependent variable, and with the known time point as the independent variable, the exponential decay curve is fitted using the nonlinear least squares method to obtain a first-order kinetic decay model.

[0040] In one specific embodiment of this example, in step a, for each time-specific core biomarker, quantitative data at five known postmortem time points (1 hour, 12 hours, 24 hours, 48 ​​hours, and 72 hours) are extracted from a multi-omics time series matrix. Each known time point contains no fewer than six biological replicates, and the arithmetic mean of each replicate is taken as the representative concentration value for that time point. The concentration values ​​at the five known time points are arranged in chronological order to form the relative concentration sequence of the time-specific core biomarker. This sequence fully records the concentration change trajectory of the biomarker from the early to late postmortem period (1 hour to 72 hours), providing basic data for subsequent kinetic modeling.

[0041] In one specific embodiment of this example, in step b, for each time-specific core biomarker, its concentration at 1 hour post-mortem (the baseline sampling time point) is set as the initial concentration, serving as the starting point of the curve. The standardized concentrations at known post-mortem time points (12 hours, 24 hours, 48 ​​hours, and 72 hours) are used as the dependent variable of the curve, and the corresponding storage time is used as the independent variable. A nonlinear least squares method is employed for curve fitting. The concentration exhibits an exponential decay trend over time, converging towards a background baseline value. The model parameters include three parameters to be fitted: initial concentration, time decay rate, and background baseline value.

[0042] The specific fitting method is as follows: For each time-specific core biomarker, its concentration data at five known time points (1 hour, 12 hours, 24 hours, 48 ​​hours, and 72 hours) are used for iterative optimization using a nonlinear least squares method. The initial concentration is set to the measured value at the 1-hour time point; the initial time decay rate is set to an empirical value (e.g., 0.1); and the initial background baseline value is set to 10% of the measured value at the 72-hour time point. The goal of the iterative optimization is to minimize the sum of squared residuals between the predicted and measured values. Convergence is determined when the parameter change rate is less than one-thousandth or the number of iterations exceeds 1000.

[0043] After fitting, the coefficient of determination (COD) is calculated to evaluate the goodness of fit. A COD of at least 0.85 is required; otherwise, the time-specific core biomarker is considered non-compliant and discarded. This embodiment verifies that the CODs of all 12 time-specific core biomarkers are greater than 0.90, with an average COD of 0.95, indicating that the model can accurately describe the concentration decay pattern of post-slaughter biomarkers in Kazakh mutton. Furthermore, a normality test is performed on all fitting residuals, and the p-values ​​are all greater than 0.05, indicating the absence of systematic bias and the scientific validity of the constructed first-order kinetic decay model.

[0044] Through the above fitting process, the individualized time decay rate of each time-specific core biomarker is obtained. The larger the decay rate value, the higher the sensitivity of the biomarker to time changes, and the faster its contribution in the dynamic confidence weight model decays with the extension of storage time.

[0045] Furthermore, the first-order dynamic decay model satisfies the following formula: in, For the first A time-specific core biomarker at a known time point Standardized concentration, For the first Several time-specific core biomarkers at the baseline sampling time point The initial concentration, For the first Background baseline values ​​for the concentration detection of time-specific core biomarkers. It is a natural constant. For the first The time decay rate of a time-specific core biomarker.

[0046] Experimental results showed that the goodness of fit for all 12 time-specific core biomarkers was significantly higher than the threshold, with an average coefficient of determination R0. 2 =0.952. The specific fitting parameters and validation results are shown in Table 1 below: Table 1 (Note: The concentration-related parameters in the table have been standardized; only key fitting indicators are shown.) The table shows that: all markers R 2 All values ​​were above 0.90, with metabolomics markers showing particularly high average R values ​​due to their well-defined biochemical pathways. 2 The result was as high as 0.976, proving that the first-order kinetic decay model can accurately describe the decay law of post-slaughter markers in Kazakh mutton. The Shapiro-Wilk normality test was performed on all fitted residuals, and the p-values ​​were all >0.05, indicating that the first-order kinetic decay model has no systematic bias and is scientifically effective.

[0047] Furthermore, the aforementioned omics pathways include proteomics pathways, metabolomics pathways, and volatile substance metabolomics pathways; In step 3 above, obtaining the support score of each omics channel for the candidate time set based on the individualized time decay rate and the pre-built dynamic confidence weight model may include the following steps: Step 3.1: Input the individualized time decay rate into the pre-built dynamic confidence weight model to obtain the dynamic confidence weight; Step 3.2: Based on the dynamic confidence weight, determine the support score for each candidate time point of the proteomics channel, the metabolomics channel, and the volatile metabolomics channel; Step 3.3: Based on the support score of each candidate time point, determine the support score of each omics channel for the candidate time set.

[0048] In step 3.1 above, the pre-construction process of the dynamic confidence weight model includes: Step 3.1.1: Determine the basic weights of the omics channels based on the individualized time decay rate; Step 3.1.2: Based on the weights of each unknown storage time point to be tested, determine the species correction factor and temperature correction factor; Step 3.1.3: Construct a dynamic confidence weight model based on the basic weights of the omics channels, the species correction factor, and the temperature correction factor; The dynamic confidence weight model satisfies the following formula: in, For the first A time-specific core biomarker at an unknown storage time. Dynamic confidence weights, The storage time to be tested is unknown. The baseline weights for each time-specific core biomarker, The time delay is the time difference between the unknown storage time to be tested and the baseline sampling time. For the first The time decay rate of a time-specific core biomarker As the basic weights for omics channels, As a species correction factor, This is a temperature correction factor; When the unknown storage time to be tested Equal to candidate time At that time, the dynamic confidence weight calculated by the dynamic confidence weight model is the first... A time-specific core biomarker at candidate time The confidence weight.

[0049] In one specific embodiment of this example, step 3.1 involves inputting the individualized time decay rate of each time-specific core biomarker obtained from step b into a pre-constructed dynamic confidence weight model. This dynamic confidence weight model comprehensively considers correction factors across multiple dimensions, including: the biomarker's base weight (for example, allocated according to omics stability, with a baseline value of 1.0 for the proteome, 0.9 for the metabolome, and 0.8 for the volatile metabolome), time delay (the time difference between the unknown storage time of the sample and the baseline time point of 1 hour post-mortem), omics channel weight, species correction factor (1.15 for Kazakh sheep and 1.0 for common sheep), and temperature correction factor (based on 4°C, with approximately 2% correction for every 1°C deviation).

[0050] The core mechanism of the dynamic confidence weight model is that the confidence weight of a biomarker decreases exponentially with storage time, reflecting the objective law that the information entropy of a biomarker gradually decreases over time. Specifically, for each storage time point (12 hours, 24 hours, 48 ​​hours, 72 hours) in the storage time set, it is assumed that the storage time of the sample to be tested is equal to that candidate time point. This is then substituted into the dynamic confidence weight model to calculate the dynamic confidence weight of each time-specific core biomarker under this assumption. This weight value ranges from 0 to 1.5; a higher weight indicates a higher reliability of the time-specific core biomarker at the current time point.

[0051] Then, based on the dynamic confidence weights, the time-specific core biomarkers can be screened for validity: when the weight is greater than or equal to 0.6, the biomarker is in the high confidence interval, and the result is valid; when the weight is between 0.3 and 0.6, it is in the medium confidence interval, and the result needs cross-validation; when the weight is less than 0.3, it is in the low confidence interval, the result is unreliable, and it should be removed. This can be expressed as follows: when When ≥0.6: the first The time-specific core biomarkers are in the high confidence interval, indicating that the judgment result is valid; When 0.3≤ When <0.6: the first The time-specific core biomarkers are in the medium confidence interval, and the judgment results need to be cross-validated. when When <0.3: the first If a time-specific core biomarker is in the low confidence range, the judgment result is unreliable and it should be removed.

[0052] For unknown storage time to be tested The solution employs a candidate time traversal assignment method: the candidate sampling time set... (exclude Each candidate value (because actual testing cannot be completed within 1 hour) Assume in turn Substitute it into the above The calculation formula yields each candidate time. Corresponding dynamic confidence weights of core biomarkers This is used for subsequent support score calculation. The principle of the candidate time traversal assignment method is as follows: Figure 2 As shown; In one specific embodiment of this example, in step 3.2 above, the candidate time set is... For each candidate time point (12 hours, 24 hours, 48 ​​hours, and 72 hours), support scores were independently calculated for the three omics channels: proteomics, metabolomics, and volatile metabolomics. The support score for each channel was calculated by summing the products of the dynamic confidence weights of all time-specific core biomarkers within that channel and a binary indicator function, and then dividing by the total number of time-specific core biomarkers for that channel. Support Score The formula is ,in A subset of core markers for omics pathways; A binary indicator function (based on the current detection value of the sample, inferring whether the sample might be at time point). Binary judgment; if the first Time-specific core biomarkers at candidate sampling time points If the concentration is higher than the instrument's limit of quantitation, the value is 1; if it is not detected or is lower than the limit of quantitation, the value is 0. A subset of core markers for omics pathways The total number of markers (denominator normalized to make channels comparable). This allows for a preliminary determination of the result. That is, the time with the highest support score is selected as the preliminary result for that channel.

[0053] The binary indicator function is based on the current detection value of the sample: if the detection concentration of the time-specific core biomarker in the sample is higher than the instrument's limit of quantitation, the value is 1; if it is not detected or is lower than the limit of quantitation, the value is 0. The purpose of denominator normalization is to eliminate the influence of differences in the number of biomarkers between different channels, making the support scores of each channel comparable.

[0054] Taking the proteomics channel as an example, this channel contains seven time-specific core biomarkers (such as troponin T, enolase 1, and lactate dehydrogenase). For the candidate time point of 24 hours, the dynamic confidence weight of each protein biomarker is first calculated under the assumption of a 24-hour storage time. Then, the value of the binary indicator function (1 for detection, 0 for non-detection) is determined based on the actual detection status of the biomarker in the sample. The weight of each biomarker is multiplied by the indicator function value, summed, and then divided by 7 to obtain the support score of the proteomics channel for the 24-hour candidate time point. Similarly, the support scores of the channel for 12 hours, 48 ​​hours, and 72 hours are calculated respectively.

[0055] The metabolomics channels (containing three time-specific core biomarkers: lactate, hypoxanthine, and inosine) and the volatile matter metabolomics channels (containing two time-specific core biomarkers: hexanal and 1-octen-3-ol) were calculated using the same logic to obtain the support scores of each channel for each candidate time point.

[0056] In one specific implementation of this embodiment, in step 3.3 above, for each omics channel, its support scores for four candidate time points (12 hours, 24 hours, 48 ​​hours, and 72 hours) are compared, and the candidate time point with the highest support score is selected as the preliminary judgment result of the channel.

[0057] Specifically, the proteomics channel calculates four support scores for each of the four candidate time points, and the time point corresponding to the highest support score is taken as the preliminary judgment result of the proteomics channel. Similarly, the metabolomics channel and the volatile metabolomics channel take the time point with the highest support score as their respective preliminary judgment results.

[0058] At this point, each of the three omics channels outputs a preliminary judgment result (each result corresponds to one of 12 hours, 24 hours, 48 ​​hours, or 72 hours). These three preliminary judgment results will serve as input for the subsequent step 3.4 (calculating the cross-omics conflict index) to quantify the degree of consistency among the judgment results of the three omics channels.

[0059] For example, if the support score for the proteomics channel is highest at 24 hours, the metabolomics channel is highest at 24 hours, and the volatile metabolomics channel is highest at 12 hours, then the preliminary judgment results for the three channels are: 24 hours, 24 hours, and 12 hours, respectively. This combination of results will be used to calculate the cross-omics conflict index to determine whether there is conflict of evidence and the severity of the conflict.

[0060] The calculation of the cross-omics conflict index based on the preliminary judgment results of each omics channel in step 3 above may include the following steps: Step 3.1B: Convert the preliminary judgment results of the proteomics channel, metabolomics channel and volatile substance metabolomics channel in the omics channel into hours respectively to obtain the preliminary judgment results of the hours of each omics channel; Step 3.2B: Based on the preliminary determination results of the number of hours of the proteomics channel, the metabolomics channel and the volatile substance metabolomics channel, calculate the cross-omics conflict index according to the pre-constructed cross-omics conflict index model; The cross-omics conflict index model satisfies the following formula: Wherein, β is the cross-omics conflict index. This is a preliminary result for determining the number of hours in the proteomics channel. This is a preliminary result for determining the number of hours in the metabolomics pathway. Preliminary results for determining the number of hours in the metabolomics pathways of volatile substances. The maximum time span is preset. The preliminary judgment results of the three omics (i.e., candidate sampling time points) ~ (corresponding number of hours) (Maximum time span). If β ≥ 0.7, there is a high degree of consistency among omics, and the majority decision result is output directly; if 0.5 ≤ β < 0.7, there is a slight conflict, and the weighted average result is used; if β < 0.5, there is a significant conflict, and the arbitration mechanism is initiated.

[0061] In one specific implementation of this embodiment, in step 3.1B, since the preliminary judgment results of the three omics channels obtained in step 3.3 are all discrete time points (12 hours, 24 hours, 48 ​​hours, or 72 hours) in the candidate time set, in order to facilitate the subsequent quantitative calculation of the conflict index, these discrete time points are first uniformly converted into the corresponding number of hours.

[0062] The specific conversion rules are as follows: 12 hours is converted to hour 12, 24 hours to hour 24, 48 hours to hour 48, and 72 hours to hour 72. After conversion, the preliminary determination results of proteomics channels correspond to one hour value, denoted as proteomics time; the preliminary determination results of metabolomics channels correspond to one hour value, denoted as metabolomics time; and the preliminary determination results of volatile substance metabolomics channels correspond to one hour value, denoted as volatile substance metabolomics time.

[0063] For example, if the preliminary determination result for the proteomics channel is 24 hours, the preliminary determination result for the metabolomics channel is 24 hours, and the preliminary determination result for the volatile metabolomics channel is 12 hours, then the converted result is: proteomics time is 24 hours, metabolomics time is 24 hours, and volatile metabolomics time is 12 hours. These three hour values ​​will be used as input parameters for calculating the cross-omics conflict index in step 3.2B.

[0064] In one specific implementation of this embodiment, in step 3.2B, based on the preliminary determination results of the number of hours of the three omics channels obtained in step 3.1B, a pre-constructed cross-omics conflict index model is input to calculate a cross-omics conflict index value between 0 and 1. This cross-omics conflict index is used to quantify the degree of consistency or conflict among the determination results of the three omics channels.

[0065] The calculation logic of the cross-omics conflict index model is as follows: First, calculate the absolute difference between each pair of the three omics judgment results, namely the absolute difference between the proteomics time and the metabolomics time, the absolute difference between the proteomics time and the volatile metabolomics time, and the absolute difference between the metabolomics time and the volatile metabolomics time; then, sum these three absolute differences; finally, compare the sum with the maximum possible time span (i.e., the maximum time span of 72 hours multiplied by 2) to obtain the cross-omics conflict index.

[0066] The closer the cross-omics conflict index is to 1, the more consistent the judgment results of the three omics channels are; the closer the cross-omics conflict index is to 0, the more dispersed the judgment results of the three omics channels are, and the more significant the conflict is.

[0067] Based on the numerical range of the cross-omics conflict index, the consistency among omics is graded and determined: When the cross-omics conflict index is greater than or equal to 0.7, it is considered that there is a high degree of consistency among omics. At this time, the judgment results of the three omics channels are basically consistent, and the majority judgment result can be directly used as the final output without the need to initiate an arbitration mechanism.

[0068] When the cross-omics conflict index is between 0.5 and 0.7, it is considered a mild inter-omics conflict. At this point, there is some disagreement among the three omics pathways, but the degree of disagreement is within an acceptable range, and a weighted average method can be used to output the results.

[0069] When the cross-omics conflict index is below 0.5, it is judged as a significant conflict between omics. At this time, there is a serious discrepancy in the judgment results of the three omics channels. Simple weighted fusion cannot effectively resolve the contradiction of evidence. It is necessary to initiate an arbitration mechanism (i.e., steps 3.1A to 3.4A) to redistribute the weights of the omics channels in order to achieve intelligent resolution of the conflict.

[0070] For example, if the proteomics timeframe is 24 hours, the metabolomics timeframe is 24 hours, and the volatile matter metabolomics timeframe is 12 hours, then the pairwise absolute differences are: 0 for protein and metabolomics, 12 for protein and volatile matter metabolomics, and 12 for metabolomics and volatile matter metabolomics, with a sum of 24. Using a maximum time span of 72 hours, the conflict index is 1 minus 24 divided by 144 (i.e., 2 multiplied by 72), resulting in a cross-omics conflict index of approximately 0.83. This value is greater than 0.7, indicating a high degree of consistency between omics, eliminating the need for arbitration and allowing direct output of the majority decision (24 hours).

[0071] In another example, if the proteomics time is 12, the metabolomics time is 48, and the volatile matter metabolomics time is 72, then the pairwise absolute differences are: 36 between protein and metabolomics, 60 between protein and volatile matter metabolomics, and 24 between metabolomics and volatile matter metabolomics, with a sum of 120. The calculated cross-omics conflict index is approximately 0.17. This value is below 0.5, indicating a significant inter-omics conflict, requiring the initiation of an arbitration mechanism to adjust the weights of the omics pathways.

[0072] By calculating and classifying the cross-omics conflict index, this method can quantitatively assess the consistency between multi-omics evidence and automatically decide whether to initiate an arbitration mechanism based on the degree of conflict, thus realizing intelligent identification and graded processing of multi-source evidence conflicts.

[0073] In step 3 above, such as Figure 3 As shown, when the cross-omics conflict index is lower than a preset threshold, a conflict arbitration mechanism is initiated to determine the weight of each omics channel, which may include the following steps: Step 3.1A: When the cross-omics conflict index is lower than a preset threshold, determine the standard deviation of the determination results of each time-specific core biomarker within each omics channel; Step 3.2A: Determine the dispersion of time-specific core biomarkers within each omics channel based on the standard deviation. Step 3.3A: Based on the dispersion of each time-specific core biomarker within each omics channel, determine the internal consistency coefficient of each omics channel; Step 3.4A: Adjust the basic weights of the omics channels according to the internal consistency coefficient to obtain the adjusted weights of each omics channel.

[0074] In one specific implementation of this embodiment, in step 3.1A, it is first determined whether the cross-omics conflict index is lower than a preset threshold of 0.5. If the conflict index is lower than 0.5, it indicates that there is a significant conflict between the preliminary judgment results of the three omics channels, and at this time, the arbitration mechanism is initiated. The first step of the arbitration mechanism is to calculate the standard deviation of the judgment results of each time-specific core biomarker within each omics channel.

[0075] Specifically, the proteomics channel contains seven time-specific core biomarkers (such as troponin T, enolase 1, and lactate dehydrogenase). Each time-specific core biomarker, based on its concentration change curve, can independently be used to infer a most likely corresponding storage time point (i.e., the known sampling time point corresponding to the peak concentration of the biomarker, or the best-matching time point inferred from its concentration value). The time points inferred from each of these seven time-specific core biomarkers are combined into a time series, and the standard deviation of this series is calculated. The standard deviation is calculated as follows: first, the arithmetic mean of the seven time points is calculated; then, the sum of squares of the deviations of each time point from the mean is calculated, divided by the degrees of freedom (7 minus 1), and finally, the square root is taken to obtain the standard deviation. This standard deviation reflects the degree of dispersion among the determination results of each time-specific core biomarker within the proteomics channel.

[0076] Similarly, for the metabolomics channel (containing three time-specific core biomarkers: lactate, hypoxanthine, and inosine) and the volatile substance metabolomics channel (containing two time-specific core biomarkers: hexanal and 1-octen-3-ol), the same method was used to calculate the standard deviation of the time-specific core biomarker determination results within each channel.

[0077] In one specific embodiment of this example, in step 3.2A, the standard deviation calculated in step 3.1A is used as a quantitative indicator to measure the dispersion of marker determination results within each omics channel. The larger the standard deviation, the more dispersed the determination results of time-specific core markers within the channel, the worse the consistency between them, and the lower the reliability of the channel; the smaller the standard deviation, the more concentrated the determination results of markers within the channel, the better the consistency between them, and the higher the reliability of the channel.

[0078] Taking the proteomics channel as an example, if the judgment results of the seven protein markers are highly concentrated around 24 hours, the calculated standard deviation is small, indicating that the channel has good internal consistency and the judgment results are reliable. Conversely, if the judgment results of the seven markers are scattered across multiple time points such as 12 hours, 24 hours, and 48 hours, the standard deviation is large, indicating that there is significant discrepancy within the channel and the judgment results are less reliable.

[0079] In one specific embodiment of this example, in step 3.3A, the dispersion (i.e., standard deviation) obtained in step 3.2A is... ), The known sampling time point corresponding to the concentration peak of the nth time-specific core biomarker. ; Preliminary channel determination results (candidate sampling time points) ), This represents the set of time-specific core biomarkers contained in the g-th omics channel; then, the internal consistency coefficient of each omics channel is calculated. , >0.8 indicates high consistency, 0.6≤ ≤0.8 is considered normal. A value less than 0.6 indicates poor internal consistency. The basic principle of internal consistency coefficient calculation is: the larger the standard deviation, the smaller the internal consistency coefficient; the smaller the standard deviation, the larger the internal consistency coefficient. The range of the internal consistency coefficient is from 0 to 1, and the closer the value is to 1, the better the internal consistency.

[0080] Specifically, when the standard deviation of the marker judgment results within a channel is close to 0, the internal consistency coefficient is close to 1, indicating that the judgment results of each marker within that channel are highly consistent. When the standard deviation is large, the internal consistency coefficient decreases accordingly. Based on the numerical range of the internal consistency coefficient, the reliability of the channel is graded: an internal consistency coefficient greater than 0.8 is considered highly consistent; an internal consistency coefficient between 0.6 and 0.8 is considered moderately consistent; and an internal consistency coefficient less than 0.6 is considered poorly consistent.

[0081] In one specific embodiment of this example, in step 3.4A, based on the internal consistency coefficients of each omics pathway determined in step 3.3A, the basic weights of the omics pathways defined in step 2.4 (1.0 for proteomics, 0.9 for metabolomics, and 0.8 for volatile metabolomics) are adjusted. The adjustment rules are as follows: When the internal consistency coefficient of a channel is greater than 0.8 (highly consistent), the channel is considered highly reliable, and its weight is increased by 50% from the original base weight. That is, the weight of the proteomics channel is adjusted to 1.5, the weight of the metabolomics channel is adjusted to 1.35, and the weight of the volatile metabolomics channel is adjusted to 1.2.

[0082] When the internal consistency coefficient of a channel is between 0.6 and 0.8 (general consistency), the channel has moderate reliability, and its weight remains unchanged. That is, the proteomics channel maintains 1.0, the metabolomics channel maintains 0.9, and the volatile metabolomics channel maintains 0.8.

[0083] When the internal consistency coefficient of a channel is lower than 0.6 (poor consistency), the reliability of that channel is low, and its weight is reduced by 50% from the original base weight. That is, the weight of the proteomics channel is adjusted to 0.5, the weight of the metabolomics channel is adjusted to 0.45, and the weight of the volatile metabolomics channel is adjusted to 0.4.

[0084] Weight of the g-th omics channel , ,in The basic weights of the omics channel g; adjustment function : >0.8 is taken as 1.5 (weight increased by 50%), 0.6≤ ≤0.8 is taken as 1.0 (unchanged). <0.6 is replaced by 0.5 (a 50% reduction); the adjusted omics channel weights will replace the original base weights and be used for calculating the comprehensive confidence score in subsequent step 4. This adjustment mechanism achieves intelligent resolution of conflicts in multi-omics evidence: when the judgment results of markers within a channel are highly consistent, its weight is increased, allowing it to play a greater role in the final judgment; when the judgment results of markers within a channel are scattered and inconsistent, its weight is reduced, weakening its impact on the final judgment. This dynamic weight adjustment strategy based on internal consistency coefficients effectively solves the problem that traditional weighted fusion methods cannot handle evidence conflicts, achieving a leap from "data fusion" to "evidence reasoning".

[0085] For example, if the internal consistency coefficient of the proteome channel is 0.85 (highly consistent), the internal consistency coefficient of the metabolome channel is 0.65 (moderately consistent), and the internal consistency coefficient of the volatile matter metabolome channel is 0.55 (poorly consistent), then the adjusted channel weights are: proteome 1.5, metabolome 0.9, and volatile matter metabolome 0.4. These adjusted omics channel weights will be used to recalculate the overall confidence score, ensuring that channels with high reliability dominate the final determination.

[0086] In step 4 above, determining the evaluation result of the post-slaughter storage time of mutton based on the weights of each omics channel and the support score may include the following steps: Step 4.1: Calculate the comprehensive confidence score for each candidate time based on the weights of each omics channel and the support score; Step 4.2: After normalizing the comprehensive confidence score, combine the cross-omics conflict index and probability value to perform a three-level confidence rating, and determine the evaluation result of the post-slaughter storage time of mutton with quantitative confidence. The comprehensive confidence score satisfies the following formula: in, To calculate the overall confidence score, For omics channel, As a time-specific core biomarker, It is the set of time-specific core biomarkers contained in the g-th omics channel. For the first A time-specific core biomarker at candidate time Confidence weights; For the first A time-specific core biomarker at candidate time The indicator function value, when the first A time-specific core biomarker at candidate time When detected, the indicator function value is 1; otherwise, it is 0. For the g-th omics channel weight, After screening ( The total number of time-specific core biomarkers with a value ≥ 0.3.

[0087] In one specific implementation of this embodiment, in step 4.1 above, based on the weights of each omics channel adjusted in step 3.4A (or the basic weights used when the arbitration mechanism is not activated), and the dynamic confidence weights and binary indicator functions of each time-specific core biomarker calculated in step 3.2, a comprehensive confidence score is calculated for each candidate time point (12 hours, 24 hours, 48 ​​hours, 72 hours) in the candidate time set.

[0088] The overall confidence score is calculated as follows: First, all effective time-specific core biomarkers within all omics channels (proteomics, metabolomics, and volatile metabolomics) are cumulatively summed. The contribution value of each time-specific core biomarker is calculated as follows: the dynamic confidence weight of the time-specific core biomarker at the assumed current candidate time point, multiplied by the binary indicator function value of the time-specific core biomarker in the sample (1 for detection, 0 for non-detection), and then multiplied by the adjusted weight of the omics channel to which the biomarker belongs. The above products are then summed over all effective time-specific core biomarkers to obtain the numerator.

[0089] The effective time-specific core biomarker is defined as a biomarker with a weight greater than or equal to 0.3 in the dynamic confidence weight screening in step 2.4. Biomarkers with a weight lower than 0.3 are eliminated due to their low confidence and are not included in the calculation of the overall confidence score.

[0090] Then, the summation result is divided by the total number of valid time-specific core biomarkers (i.e., the number of time-specific core biomarkers involved in the calculation) to obtain the overall confidence score for the candidate time point. This score reflects the overall support of all valid multi-omics evidence for the decision at the current hypothetical time point. The higher the overall confidence score, the more likely the storage time of the sample belongs to the candidate time point.

[0091] Following the above method, the comprehensive confidence scores for the four candidate time points (12 hours, 24 hours, 48 ​​hours, and 72 hours) were calculated respectively, resulting in four comprehensive confidence score values.

[0092] In one specific embodiment of this example, in step 4.2 above, the comprehensive confidence scores of the four candidate time points calculated in step 4.1 are first converted into a probability distribution form. The specific method for probability normalization is as follows: the comprehensive confidence score of each candidate time point is divided by the sum of the comprehensive confidence scores of the four candidate time points to obtain the normalized probability value of that candidate time point. That is After normalization, the sum of the probability values ​​for all candidate time points is 1, which is... Each probability value, ranging from 0 to 1, represents the likelihood that the storage time of the sample belongs to that candidate time point. The higher the probability value, the higher the confidence level in attributing it to that time point.

[0093] In step 4.2 above, after normalizing the overall confidence score, a three-level confidence rating is performed by combining the cross-omics conflict index and the probability value to determine the assessment result of the post-slaughter storage time of mutton with quantitative confidence. This may include the following steps: Step 4.2.1: Sort the normalized probability values ​​of each candidate time point from largest to smallest, and denot them as the first probability value, the second probability value, and so on. Step 4.2.2: Based on the normalized probability value, output the evaluation result according to the following three-level judgment rules: Step 4.2.3: First-level determination is a clear determination: When the first probability value is greater than the first probability threshold, and the difference between the first probability value and the second probability value is greater than the preset difference threshold, it is determined as the candidate time point corresponding to the first probability value, and the post-slaughter storage time of the determined time point is output; Step 4.2.4: Secondary determination is interval determination: When the first probability value is greater than the second probability threshold, and the difference between the first probability value and the second probability value is less than or equal to the preset difference threshold, it is determined that the candidate time point corresponding to the first probability value and the candidate time point corresponding to the second probability value form an adjacent time interval, and the post-mortem storage time of the interval range is output. Step 4.2.5: Level 3 determination is undeterminable: When the first probability value is less than the third probability threshold, it is determined that a valid conclusion cannot be drawn, and the result of undeterminable is output and a prompt for resampling and detection is given.

[0094] In one specific embodiment of this example, in step 4.2.1, the normalized probability values ​​of the four candidate time points (12 hours, 24 hours, 48 ​​hours, and 72 hours) calculated in step 4.1 are arranged in descending order of numerical value. After arrangement, the maximum value is recorded as the first probability value max[P( The second largest value is denoted as the second probability value. The third largest value is denoted as the third probability value, and the smallest value is denoted as the fourth probability value. Each probability value corresponds to a specific candidate time point. Through sorting, the candidate time points with the highest and second highest probabilities and their corresponding probability value differences can be clearly identified, providing a basis for subsequent three-level judgments.

[0095] In one specific embodiment of this example, in step 4.2.2, based on the first probability value, the second probability value, and their corresponding candidate time points obtained in step 4.2.1, the evaluation results are output in a tiered manner according to a preset three-level judgment rule. The three-level judgment rule includes three scenarios: explicit judgment, interval judgment, and indeterminate judgment, each corresponding to different probability conditions and output formats.

[0096] In one specific implementation of this embodiment, in step 4.2.3, the first probability threshold is 0.6, and the preset difference threshold can be set to 0.15. When the first probability value is greater than 0.6, and the difference between the first probability value and the second probability value is greater than 0.15, it indicates that the advantage of the candidate time point with the highest probability is extremely significant, the data is highly persuasive, and it can accurately pinpoint a specific time point. At this time, a first-level explicit judgment is performed, and the candidate time point corresponding to the first probability value is output as the final evaluation result. An example of the output format is: "Post-slaughter storage time is 24 hours". This result is the most accurate and reliable evaluation conclusion. It can be expressed as: Existence exists. Make max[P ( )]>0.6, and max[P ( )]- >0.15, output "Post-slaughter storage time is..." (High confidence level)

[0097] In one specific implementation of this embodiment, in step 4.2.4, the second probability threshold is 0.5, and the preset difference threshold is set to 0.15. When the first probability value is greater than 0.5, but the difference between the first probability value and the second probability value is less than or equal to 0.15, it indicates that although the candidate time point with the highest probability has a certain advantage, the difference between it and the second highest probability time point is very small, making it impossible to accurately pinpoint a specific time point. At this time, a secondary interval judgment is performed, and the adjacent time interval formed by the candidate time point corresponding to the first probability value and the candidate time point corresponding to the second probability value is output as the final evaluation result. It should be noted that the candidate time points corresponding to the first probability value and the second probability value must be adjacent on the time axis (e.g., 24 hours and 48 hours). If they are not adjacent, further verification is required. An example of the output format is: "Post-slaughter storage time is between 24 hours and 48 hours". This result is an interval judgment, which ensures the rigor of the evaluation and avoids the risk of misjudgment caused by reporting a single point. It can be expressed as: There exists Make max[P ( )]>0.5, but max[P ( )]- ≤0.15, output "Post-slaughter storage time is..." to Interval (medium confidence level) , (These are the two adjacent candidate time points with the highest probability).

[0098] In one specific implementation of this embodiment, the third probability threshold is set to 0.4. When the first probability value is less than 0.4, it indicates that the normalized probability values ​​of all candidate time points are at a low level, the data support is insufficient, and a valid evaluation conclusion cannot be drawn. At this time, a level 3 decision cannot be made, and instead of outputting the specific storage time, an "undeterminable" result is output, suggesting resampling or other auxiliary methods. An example of the output format is: "Undeterminable, resampling is recommended." This decision effectively avoids making incorrect judgments when the data quality is insufficient, ensuring the reliability and rigor of the evaluation method. It can be expressed as: max[P ( If the value is less than 0.4, the output cannot be determined, and it is recommended to supplement with fresh samples for testing or use other auxiliary methods.

[0099] Finally, based on the cross-omics conflict index β and the maximum probability max[P ( The credibility rating of the judgment results can be specifically shown in Table 2 below: Table 2 For example, if the normalized probability values ​​of a sample to be tested are ranked as follows: 24 hours (0.78), 12 hours (0.12), 48 hours (0.06), 72 hours (0.04), then the first probability value is 0.78 (corresponding to 24 hours), and the second probability value is 0.12 (corresponding to 12 hours). The first probability value of 0.78 is greater than 0.6, and the difference between the first probability value and the second probability value of 0.66 is greater than 0.15. Therefore, a first-level explicit judgment is performed, and the output is "Post-slaughter storage time is 24 hours".

[0100] In another example, if the normalized probability values ​​are sorted as follows: 24 hours (0.52), 48 hours (0.40), 12 hours (0.05), 72 hours (0.03), then the first probability value is 0.52 (corresponding to 24 hours), and the second probability value is 0.40 (corresponding to 48 hours). The first probability value of 0.52 is greater than 0.5, but the difference between it and the second probability value of 0.12 is less than or equal to 0.15. Therefore, a second-level interval determination is performed, and the output is "the post-slaughter storage time is in the range of 24 hours to 48 hours".

[0101] For another example, if the normalized probability values ​​are sorted as follows: 24 hours (0.35), 12 hours (0.30), 48 hours (0.20), 72 hours (0.15), then the first probability value of 0.35 is less than 0.4, so the third-level test cannot be performed, and the output is "Cannot be determined, it is recommended to resample and detect".

[0102] Through the above three-level judgment rules, this method can output evaluation results with different precisions based on different characteristics of probability distribution, providing effective information to the greatest extent while ensuring accuracy, and meeting the differentiated needs of different application scenarios for the reliability of evaluation results.

[0103] To verify the effectiveness of this method, 45 blind samples of Kazakh sheep (with known actual storage time) were tested using this method. The test results are shown in Table 3 below: Table 3 As shown in Table 3, the test results indicate that the overall accuracy of this method for 45 blind samples reached 91.1%, with accuracy rates of 91.7% at 24h, 48h, and 72h time points, and 88.9% at 12h time point. The average confidence level was 0.77, reaching level B (medium confidence) or higher, indicating that this method has good evaluation performance and stability at different storage time points.

[0104] The results of comparing this method with the traditional static threshold method are shown in Table 4 below: Table 4 As shown in Table 4, compared with the traditional static threshold method, this method improves accuracy by 18.3%, reduces false positive rate by 67.3%, and improves temporal resolution from 1 day (coarse) to supporting evaluation at discrete time points with 12-hour intervals (fine), achieving a 2-fold improvement in temporal resolution. Furthermore, this method pioneers a multi-omics data fusion evaluation mechanism, effectively solving the problem that traditional methods cannot handle conflicts in multi-source evidence.

[0105] In summary, the method for assessing post-slaughter storage time of mutton based on multi-omics data provided in this application significantly improves the accuracy and temporal resolution of post-slaughter storage time assessment by dynamically quantifying the time decay characteristics of biomarkers, intelligently resolving conflicts in multi-omics evidence, and outputting assessment results with credibility grading. This meets the industry's needs for precise traceability and quality grading of high-end mutton products.

[0106] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0107] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0108] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0109] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0110] In addition, each functional unit in the various embodiments of this application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0111] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause the device automatic test line to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0112] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0113] The features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method or device embodiments.

[0114] The above description is merely an 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.

Claims

1. A method for evaluating the post-slaughter storage time of mutton based on multi-omics data, characterized in that, The method includes: Samples of the longissimus dorsi muscle were collected from Kazakh sheep at multiple known time points after slaughter. Quantitative data of proteins, metabolites, and volatile compounds were obtained using proteomics, metabolomics, and volatile metabolomics technologies, respectively. After standardization, a multi-omics time series matrix was constructed. Some of the known time points were used as a candidate time set. For the multi-omics time series matrix, a pre-constructed nested strategy of two-level comparison and exclusion screening is adopted to screen out multiple time-specific core biomarkers; the individualized time decay rate of each time-specific core biomarker is determined by a pre-constructed first-level kinetic decay model. Based on the individualized time decay rate and the pre-built dynamic confidence weight model, the support score of each omics channel to the candidate time set is obtained; based on the support score, the preliminary judgment result of each omics channel is determined; based on the preliminary judgment result of each omics channel, the cross-omics conflict index is calculated; when the cross-omics conflict index is lower than a preset threshold, the conflict arbitration mechanism is activated to determine the weight of each omics channel. The evaluation results of the post-slaughter storage time of mutton are determined based on the weights of each omics channel and the support scores.

2. The method according to claim 1, characterized in that, For the multi-omics time series matrix, a pre-constructed nested strategy of two-level comparison and exclusion screening is used to screen out multiple time-specific core biomarkers, including: The first specific difference between 24-hour and 1-hour quantitative data in the multi-omics time series matrix is ​​determined, and quantitative data with a first specific difference greater than a preset first specific difference threshold are selected as the first specific core biomarker set. The second specific difference between 72-hour and 1-hour quantitative data in the multi-omics time series matrix, and the third specific difference between 72-hour and 24-hour quantitative data in the multi-omics time series matrix are determined. Quantitative data satisfying both the second specific difference and the third specific difference being greater than a preset second specific difference threshold are selected as the second specific core biomarker set. Common biomarkers between the first and second specific core biomarker sets are removed, and the remaining biomarkers are merged to obtain multiple time-specific core biomarkers. The multiple time-specific core biomarkers include: proteomics biomarkers, metabolomics biomarkers, and volatile substance metabolomics biomarkers.

3. The method according to claim 1, characterized in that, The pre-construction process of the first-order dynamic decay model includes: Based on the multiple time-specific core biomarkers, the relative concentration sequence of each time-specific core biomarker was determined; For each time-specific core biomarker, starting from the initial concentration at the obtained baseline sampling time point, and using the standardized concentration at the obtained known time point as the dependent variable, and with the known time point as the independent variable, a nonlinear least squares method is used to fit the exponential decay curve to obtain a first-order kinetic decay model.

4. The method according to claim 3, characterized in that, The first-order dynamic decay model satisfies the following formula: in, For the first A time-specific core biomarker at a known time point Standardized concentration, For the first Several time-specific core biomarkers at the baseline sampling time point The initial concentration, For the first Background baseline values ​​for the concentration detection of time-specific core biomarkers. It is a natural constant. For the first The time decay rate of a time-specific core biomarker.

5. The method according to claim 1, characterized in that, The aforementioned omics pathways include proteomics pathways, metabolomics pathways, and volatile substance metabolomics pathways; The individualized time decay rate and the pre-constructed dynamic confidence weight model are used to obtain the support score of each omics channel for the candidate time set, including: The individualized time decay rate is input into a pre-built dynamic confidence weight model to obtain the dynamic confidence weight. Based on the dynamic confidence weights, the support scores for each candidate time point are determined for the proteomics channel, the metabolomics channel, and the volatile metabolomics channel; based on the support scores for each candidate time point, the support scores for each omics channel to the candidate time set are determined.

6. The method according to claim 5, characterized in that, The pre-construction process of the dynamic confidence weight model includes: Based on the individualized time decay rate, the basic weights of the omics channels are determined; Based on the weights of each unknown storage time point to be tested, species correction factors and temperature correction factors are determined; Based on the basic weights of the omics channels, the species correction factor, and the temperature correction factor, a dynamic confidence weight model is constructed. The dynamic confidence weight model satisfies the following formula: in, For the first A time-specific core biomarker at an unknown storage time. Dynamic confidence weights, The storage time to be tested is unknown. The baseline weights for each time-specific core biomarker; The time delay is the time difference between the unknown storage time to be tested and the baseline sampling time. For the first The time decay rate of a time-specific core biomarker As the basic weights for omics channels, As a species correction factor, This is a temperature correction factor; When the unknown storage time to be tested Equal to candidate time At that time, the dynamic confidence weight calculated by the dynamic confidence weight model is the first... A time-specific core biomarker at candidate time Confidence weight ( ).

7. The method according to any one of claims 1-6, characterized in that, The determination of the post-slaughter storage time assessment result of mutton based on the weights of each omics channel and the support score includes: Based on the weights of each omics channel and the support scores, the overall confidence score for each candidate time is calculated. After probability normalization of the comprehensive confidence score, a three-level confidence rating is performed by combining the cross-omics conflict index and the probability value to determine the evaluation result of the post-slaughter storage time of mutton with quantitative confidence. The comprehensive confidence score satisfies the following formula: in, To calculate the overall confidence score, For omics channel, As a time-specific core biomarker, It is the set of time-specific core biomarkers contained in the g-th omics channel. For the first A time-specific core biomarker at candidate time Confidence weights; For the first A time-specific core biomarker at candidate time The indicator function value, when the first A time-specific core biomarker at candidate time When detected, the indicator function value is 1; otherwise, it is 0. For the g-th omics channel weight, After screening ( The total number of time-specific core biomarkers with a value ≥ 0.

3.

8. The method according to claim 1, characterized in that, When the cross-omics conflict index is lower than a preset threshold, a conflict arbitration mechanism is initiated to determine the weights of each omics channel, including: When the cross-omics conflict index is lower than a preset threshold, the standard deviation of the determination results of each time-specific core biomarker in each omics channel is determined. Based on the standard deviation, the dispersion of time-specific core biomarkers within each omics channel is determined; Based on the degree of dispersion of time-specific core biomarkers within each omics channel, the internal consistency coefficient of each omics channel is determined. Based on the internal consistency coefficient, the basic weights of the omics channels are adjusted to obtain the adjusted weights of each omics channel.

9. The method according to claim 1, characterized in that, The step of calculating the cross-omics conflict index based on the preliminary determination results of each omics channel includes: converting the preliminary determination results of the proteomics channel, metabolomics channel and volatile substance metabolomics channel in the omics channel into hours, respectively, to obtain the preliminary determination results of the hours of each omics channel; Based on the preliminary determination results of the number of hours of the proteomics channel, the metabolomics channel, and the volatile substance metabolomics channel, the cross-omics conflict index is calculated according to the pre-constructed cross-omics conflict index model. The cross-omics conflict index model satisfies the following formula: Wherein, β is the cross-omics conflict index. This is a preliminary result for determining the number of hours in the proteomics channel. This is a preliminary result for determining the number of hours in the metabolomics pathway. Preliminary results for determining the number of hours in the metabolomics pathways of volatile substances. The maximum time span is preset.

10. The method according to claim 1, characterized in that, After normalizing the overall confidence score, a three-level confidence rating is performed by combining the cross-omics conflict index and the probability value to determine the assessment results of the post-slaughter storage time of mutton with quantitative confidence, including: The comprehensive confidence scores of each candidate time point are normalized to obtain the normalized probability value of each candidate time point; the normalized probability value is between 0 and 1, and the sum of the normalized probability values ​​of all candidate time points is 1. The normalized probability values ​​of each candidate time point are sorted from largest to smallest and denoted as the first probability value, the second probability value, and so on. Based on the normalized probability value, the evaluation result is output according to the following three-level judgment rules: The first-level determination is a clear determination: when the first probability value is greater than the first probability threshold, and the difference between the first probability value and the second probability value is greater than the preset difference threshold, it is determined as the candidate time point corresponding to the first probability value, and the post-slaughter storage time of the determined time point is output. The second-level determination is an interval determination: when the first probability value is greater than the second probability threshold, and the difference between the first probability value and the second probability value is less than or equal to the preset difference threshold, it is determined that the candidate time point corresponding to the first probability value and the candidate time point corresponding to the second probability value form an adjacent time interval, and the post-slaughter storage time within this interval is output. Level 3 determination: When the first probability value is less than the third probability threshold, it is determined that no valid conclusion can be drawn, and the result of "cannot be determined" is output and a prompt for resampling and detection is given.

Images