Crassostrea calthrop hybrid prejudgment and combined configuration method based on COI genetic distance

Abstract

The invention relates to the technical field of oyster cross breeding, in particular to a COI genetic distance-based crassostrea calthrop hybridization prejudgment and combined configuration method. The method comprises the following steps: collecting COI gene sequences of calthrop oysters and four cultured oyster samples, constructing an allelic point sliding window, and calculating genetic differences by adopting a Kimura-2P distance, so as to obtain a refined COI genetic distance index; establishing a COI heredity-hybridization advantage regression model by combining historical hybridization survival data, heredity variation data and disturbance factors, introducing a multi-dimensional threshold structure based on genetic distance, and performing regression pre-judgment of threshold logic regulation and control on the hybridization advantage; on the basis, a COI genetic distance matrix is constructed, intraspecific fertility analysis and character complementation inference are coupled, and optimization and combined configuration of cross parents are achieved. Quantitative prediction of the hybridization advantage can be achieved before hybridization, the breeding period is shortened, the hybridization test cost is reduced, and the accuracy of breeding hybridization auxiliary decision making is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of oyster hybridization breeding technology, and in particular to a method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance. Background Technology

[0002] As an important economically important shellfish species in marine aquaculture, the growth rate, survival rate, and environmental adaptability of the oyster directly affect the aquaculture benefits. Current oyster breeding methods mainly rely on phenotypic selection and multi-generation hybridization experiments. Hybrid vigor is determined by observing the growth and survival performance of hybrid offspring through long-term field or marine aquaculture. This method is time-consuming, costly, and significantly affected by environmental fluctuations, making it difficult to accurately determine the superiority of parental combinations before hybridization. Some studies have attempted to introduce molecular marker-assisted breeding, but these often focus on single or a few genetic loci, failing to systematically characterize the overall genetic differences between different oyster species or strains, resulting in insufficient stability in predicting hybrid vigor.

[0003] Existing techniques typically rely solely on simple correlation analysis based on genetic distance, neglecting the synergistic impact of environmental disturbances on hybridization survival performance through regression training in historical hybridization data. This makes it difficult for traditional regression prediction models to distinguish the contributions of genetic and non-genetic effects to survival rate improvement, resulting in low-reliability indicators for quantifying hybrid vigor. Furthermore, the lack of explicit threshold constraints on the occurrence interval of hybrid vigor easily leads to the accumulation of residual errors in predicting hybrid vigor. In addition, existing methods still rely on empirical judgment for parental combination configuration, resulting in high trial-and-error costs and low hybridization success rates. They fail to effectively screen and combine oysters with cultured parents from the perspective of trait complementarity, combining COI genetic distance and hybrid vigor rate. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and provides a method for predicting and combining hybrid vigor in oysters based on COI genetic distance.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: The first aspect of this invention provides a method for predicting and configuring hybrid vigor in Crassula spinosa based on COI genetic distance, comprising the following steps: S01: Collect and detect the base labeling sites of oysters and various farmed oysters by biological sequence detection, construct an isotope sliding window for base alignment, and use the K2P distance calculation architecture in the isotope sliding window to calculate the variation difference after the alignment of two base labeling sites on the COI genetic sequence, so as to obtain the COI genetic distance of oyster samples compared with farmed oyster samples. S02: Obtain historical hybridization survival data, historical COI genetic variation data, and perturbation factor data of oyster samples and cultured oyster samples. Estimate the propensity of the regulatory residuals that are passively treated with covariates of the perturbation factor data. Based on the propensity of the regulatory residuals, establish an observation-dependent regression relationship between historical hybridization survival data and historical COI genetic variation data to generate a COI genetic-hybridization advantage regression model. S03: Based on the threshold constraint interval of COI genetic distance driving the survival rate improvement of the hybrid group of oyster, a multidimensional threshold structure is constructed. Based on the multidimensional threshold structure, the COI genetic distance is introduced into the COI genetic-hybrid vigor regression model to perform threshold logic regulation regression prediction analysis on the hybrid vigor of oyster, and the hybrid vigor rate of oyster is obtained. S04: Based on the hybrid vigor, select candidate parents of cultured oysters and construct a COI genetic distance matrix. Combine the COI genetic distance matrix with the intraspecific fertility analysis of the threshold constraint interval to screen out the fixed parents to be hybridized. By coupling the current growth trait indicators, infer the trait complementarity bottleneck of the fixed parents to be hybridized for the oyster. Select cultured oyster individuals with the greatest trait complementarity intensity for hybridization combination pairing.

[0006] Furthermore, S01 specifically includes the following steps: Individual gene samples of oysters were collected, and multiple sequence mutation tests were performed on the COI genetic markers of the individual gene samples using a biodetection method. Explicit insertion and deletion sites were identified, and the base labeling sites of oysters were obtained and defined as a class of base labeling sites. Individual gene samples of four other farmed oyster species were extracted from the farming logs and the same bioassay process was performed to obtain the base labeling sites of different farmed oyster parents, which were defined as Class II base labeling sites. Based on the hybridization expectation of oyster, an isotopic sliding window and sliding step size for base alignment were constructed. Within the isotopic sliding window, the variation difference of COI genetic marker sequences after alignment of Class I and Class II base labeling sites was analyzed. The Kimura 2-Parameter distance calculation framework was used for cumulative calculation to obtain the genetic substitution feature value of the window. Based on the sliding step size, the above steps of mutation analysis and genetic substitution feature calculation are repeated by moving the sliding window from the sequence start point backward to the same site until the entire COI sequence is covered. The genetic distance vector is obtained, and the COI genetic distance of the oyster sample relative to the cultured oyster sample is determined based on the genetic distance vector.

[0007] Furthermore, based on the expected hybridization of the oyster *Crassostrea gigas*, an isotopic sliding window and sliding step size for base alignment are constructed. Within the isotopic sliding window, the variation differences of the COI genetic marker sequences after alignment of the first-class and second-class base labeling sites are analyzed, and the Kimura 2-Parameter distance calculation framework is used for cumulative calculation to obtain the genetic substitution feature value of the window. Specifically, this includes the following steps: Obtain hybridization expectation information of oysters, construct an isolocus sliding window for base reference alignment based on the genetic calculation granularity of COI genetic distance according to the hybridization expectation information, and preset the sliding step size according to the gene control resolution specified by the hybridization expectation information. A base sequence alignment space is constructed, and first-class base labeling sites and second-class base labeling sites are aligned in the base sequence alignment space. The Kimura 2-Parameter distance calculation architecture is introduced and coupled into the base sequence alignment space. Based on the sliding window of the equiloculation, the corresponding COI genetic marker sequence fragments between the aligned oyster samples and the cultured oyster samples are extracted from the window edge interval. A subset of COI alignment sequences under the sliding window of the equiloculation is established. Functional analysis and information entropy variation frequency weighting are performed on the base labeling sites of each window alignment interval in the subset of COI alignment sequences, and the mutation weight value of each base labeling site group is output. Based on the mutation weight values, the substitution differences of COI genetic marker sequences between the oyster sample and the cultured oyster sample are calculated by accumulating the values ​​within the current window using the Kimura 2-Parameter distance calculation architecture, and the genetic substitution feature value of the window is obtained.

[0008] Furthermore, S02 specifically includes the following steps: Obtain the breeding memo log of oyster hybridization, and extract historical hybridization survival data between oyster samples and cultured oyster samples within a preset time period, historical COI genetic variation data of each sample when different historical hybridization survival data are generated, and perturbation factor data captured during the hybridization process from the breeding memo log. Historical COI genetic variation data is defined as the independent variable and historical hybridization survival data is defined as the dependent variable. The marginal distribution between each independent variable and the dependent variable is constructed using the nonparametric distribution theorem to generate a marginal-dependent distribution model. The perturbation factor data is defined as a covariate. The probability of each historical hybridization survival data being treated by the covariate under the premise of dependence on different historical COI genetic variation data is estimated by the marginal-dependent distribution model, and the propensity of the regulatory residual of the perturbation factor data is obtained. Based on the probability of the controlled residual tendency, the perturbation factor data is introduced into the crossover distribution difference for each historical crossover survival data to calculate the tendency, and the perturbation tendency score is output. Based on the perturbation tendency score, all historical COI genetic variation data and historical hybrid survival data are jointly mapped to the [0,1] observation space for probability integral transformation, forming pseudo-observation samples of COI genetic variation-oyster hybrid survival advantage and corresponding observation dependence features. An error regression framework is introduced, and a conditional dependency function is preset based on the observation dependency characteristics. Based on the conditional dependency function, potential hybrid vigor is inferred by combining pseudo-observation samples and historical hybridization survival data in the error regression framework, thus generating the COI genetic-hybrid vigor regression model.

[0009] Furthermore, S03 specifically includes the following steps: By obtaining the threshold constraint range of COI genetic distance driving the survival rate improvement of the hybrid group of oyster, the COI genetic-hybrid vigor regression model is switched in different ranges according to the boundary of the threshold constraint range, and a multidimensional threshold structure driven by COI genetic distance is constructed. Based on the multidimensional threshold structure, the COI genetic distance is imported into the COI genetic-hybrid vigor regression model for regression prediction analysis, outputting a series of hybrid vigor regression coefficients. At the same time, the quantiles of the COI genetic distance are calculated, and the hybrid vigor regression coefficients are sorted in descending order according to the quantiles to generate a sparse regression path for hybrid vigor. After sorting in descending order, extreme coefficient samples on the sparse regression path are removed, and one or more regression coefficients of non-extreme samples are obtained, which are defined as candidate sparse regression coefficients. When the COI genetic distance is within the threshold constraint range, it is assumed that the survival rate of the hybrid group of the oyster sample increases by a positive percentage compared to the cultured oyster sample; when the COI genetic distance is not within the threshold constraint range, it is considered that the survival rate of the hybrid group does not change or decreases by a negative percentage, forming the threshold regression logic criterion for hybrid vigor. A log-likelihood algorithm is introduced. Based on the threshold regression logic criterion, the regression coefficient of each candidate coefficient is sample-likelihood fitting in the log-likelihood algorithm to generate the hybrid vigor log-likelihood value of each candidate sparse regression coefficient based on threshold logic control. The regression trend of each hybrid vigor log-likelihood value is jointly predicted according to the COI genetic distance to obtain the hybrid vigor rate of oyster.

[0010] Furthermore, S04 specifically includes the following steps: If the hybrid vigor rate is less than the preset hybrid vigor rate, the hybridization selection of the cultured oyster sample is ignored; if the hybrid vigor rate is greater than the preset hybrid vigor rate, several candidate parents to be hybridized and their corresponding actual phenotypic data are obtained for the cultured oyster sample, and the COI genetic distance matrix between the candidate parents to be hybridized and the oyster sample is constructed based on the COI genetic distance. Based on breeding experience cases, query actual phenotypic data to obtain prior distribution information on phenotypic transfer between candidate parents and different oysters. Based on the prior distribution information on phenotypic transfer, construct intraspecific gene clusters of parents. Combine the COI genetic distance matrix and threshold constraint interval to assign each parent to intraspecific gene clusters and change the corresponding fertile cluster center to generate a hybridization fertility index. Based on the hybridization fertility index, select acceptable intraspecific fixed parents to be hybridized. The current growth trait index of the fixed parent in the species to be hybridized is obtained and defined as the first growth trait index. Based on the breeding experience case search, the maximum bottleneck resistance coefficient that restricts the complementary expression of growth traits of the fixed parent in the species to be hybridized for the oyster sample is determined at the gene level. Based on the maximum bottleneck resistance coefficient, the complementary bottleneck function of the parents is constructed, and the current growth trait index of the oyster sample is obtained and defined as the second growth trait index. The complementary dominance of each first growth trait index to the hybridization genetic expression between the second growth trait index is evaluated to obtain the trait complementary coupling value. If the complementary coupling value of a trait is lower than a preset threshold, a penalty is applied to the first growth trait index corresponding to the complementary coupling value of the trait, a penalty coupling criterion term is constructed, and the penalty coupling criterion term is introduced into the complementary bottleneck function. A genetic algorithm is introduced to construct a coupled solver. The coupled solver is used to iteratively update the solutions for each first growth trait index in order to identify the optimality of the coupling for the complementarity of the spiny oyster sample after penalty coordination in the complementary bottleneck function corresponding to the fixed parents in each hybrid species, and to generate the complementarity strength of traits. Only the cultured oyster individuals corresponding to the fixed parent species of the species to be hybridized with the maximum complementary strength of the trait are extracted and paired with the spiny oyster individuals to output the hybridization combination configuration scheme of spiny oyster.

[0011] Furthermore, the method involves obtaining prior phenotypic distribution information of the candidate parents for hybridization and different oyster species by querying actual phenotypic data based on breeding experience cases, constructing intraspecific gene clusters of the parents based on the prior phenotypic distribution information, assigning each parent to an intraspecific gene cluster by combining the COI genetic distance matrix and threshold constraint intervals, and changing the corresponding fertile cluster centers to generate a hybridization fertility index. Acceptable intraspecific fixed parents for hybridization are then selected based on the hybridization fertility index. This process specifically includes the following steps: Based on big data, we can obtain breeding experience cases of candidate parents for hybridization, and obtain the prior distribution information of phenotypic transfer of candidate parents for hybridization with different oyster varieties by querying the actual phenotypic data of the breeding experience cases. Based on the prior distribution information of phenotypic transfer, intraspecific gene clusters and corresponding fertile cluster centers of different candidates for hybridization are constructed. The breeding advantage distance quantification of the COI genetic distance matrix and the threshold constraint interval is combined to appropriately allocate each candidate parent to each intraspecific gene cluster. During the allocation process, the fertile cluster centers of all parents assigned to the gene clusters of that species are replaced, and a weighted average is performed according to the normalized weights to output the current scaling stability of the fertile cluster centers. Repeat the above steps to iterate and calculate various intragene clusters. If the current scaling stability is greater than 0, stop the iteration operation immediately and finally output the squared error of the intragene cluster to which the oyster sample belongs. Based on the squared error, the hybridization combination between the oyster sample and each candidate parent to be hybridized is quantitatively evaluated based on the appropriate hybridization phenotypic performance at the advantage threshold, and a hybridization fertility index is generated. If the hybridization fertility index is greater than the maximum hybridization fertility index, the candidate parent to be hybridized corresponding to the hybridization fertility index is extracted and marked to obtain one or more fixed parents within the species to be hybridized.

[0012] This invention addresses the technical deficiencies in the prior art, and its beneficial technical effects are as follows: Base marker sites of *Crassostrea gigas* and various cultured oysters were collected and obtained through biological sequence detection. An isolocution sliding window was constructed for base alignment. Within this window, the variation difference between two aligned base marker sites on the COI genetic sequence was calculated using a K2P distance calculation framework, yielding the COI genetic distance between *Crassostrea gigas* and cultured oyster samples. Historical hybridization survival data, historical COI genetic variation data, and perturbation factor data were obtained for both *Crassostrea gigas* and cultured oyster samples. The propensity of regulatory residuals passively treated with covariates of perturbation factor data was estimated. Based on this propensity, an observation-dependent regression relationship of hybrid vigor was established between historical hybridization survival data and historical COI genetic variation data, generating the COI genetic-hybrid vigor. Regression Model: A multidimensional threshold structure was constructed based on the threshold constraint interval for improving the survival rate of the hybrid group of Crassula spinosa based on COI genetic distance. Based on the multidimensional threshold structure, COI genetic distance was introduced into the COI genetic-hybrid vigor regression model to perform threshold logic regulation regression prediction analysis on the hybrid vigor of Crassula spinosa, and the hybrid vigor rate of Crassula spinosa was obtained. According to the hybrid vigor, candidate parents of cultured oysters were selected and a COI genetic distance matrix was constructed. Combined with the intraspecific fertility analysis of COI genetic distance matrix and threshold constraint interval, the intraspecific fixed parents to be hybridized were screened. The trait complementarity bottleneck of the intraspecific fixed parents to be hybridized for Crassula spinosa was inferred by coupling the current growth trait indicators. The cultured oyster individuals with the maximum trait complementarity intensity were selected for hybridization combination pairing. This invention can construct a genetic selection basis for hybridization of Crassula spinosa based on COI genetic distance. It can quantitatively calculate COI genetic distance to predict the hybrid vigor rate of Crassula spinosa before hybridization. It can also combine fertility planning and trait complementarity strength assessment for combination configuration, thereby improving the scientific nature of the selection decision for hybridization parents of Crassula spinosa, shortening the phenotypic evaluation cycle, and improving breeding efficiency. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0014] Figure 1 The flowchart of the first method for predicting and combining hybrid vigor of Crassula spinosa based on COI genetic distance is shown. Figure 2 A flowchart of the second method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance is shown. Detailed Implementation

[0015] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0016] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0017] The first aspect of this invention provides a method for predicting and configuring hybrid vigor in *Crassostrea gigas* based on COI genetic distance, such as... Figure 1 As shown, it includes the following steps: S01: Collect and detect the base labeling sites of oysters and various farmed oysters by biological sequence detection, construct an isotope sliding window for base alignment, and use the K2P distance calculation architecture in the isotope sliding window to calculate the variation difference after the alignment of two base labeling sites on the COI genetic sequence, so as to obtain the COI genetic distance of oyster samples compared with farmed oyster samples. Among them, the full name of the K2P distance measurement architecture is Kimura 2-Parameter distance measurement architecture; S02: Obtain historical hybridization survival data, historical COI genetic variation data, and perturbation factor data of oyster samples and cultured oyster samples. Estimate the propensity of the regulatory residuals that are passively treated with covariates of the perturbation factor data. Based on the propensity of the regulatory residuals, establish an observation-dependent regression relationship between historical hybridization survival data and historical COI genetic variation data to generate a COI genetic-hybridization advantage regression model. S03: Based on the threshold constraint interval of COI genetic distance driving the survival rate improvement of the hybrid group of oyster, a multidimensional threshold structure is constructed. Based on the multidimensional threshold structure, the COI genetic distance is introduced into the COI genetic-hybrid vigor regression model to perform threshold logic regulation regression prediction analysis on the hybrid vigor of oyster, and the hybrid vigor rate of oyster is obtained. S04: Based on the hybrid vigor, select candidate parents of cultured oysters and construct a COI genetic distance matrix. Combine the COI genetic distance matrix with the intraspecific fertility analysis of the threshold constraint interval to screen out the fixed parents to be hybridized. By coupling the current growth trait indicators, infer the trait complementarity bottleneck of the fixed parents to be hybridized for the oyster. Select cultured oyster individuals with the greatest trait complementarity intensity for hybridization combination pairing.

[0018] Furthermore, S01 specifically includes the following steps: Individual gene samples of oysters were collected, and multiple sequence mutation tests were performed on the COI genetic markers of the individual gene samples using a biodetection method. Explicit insertion and deletion sites were identified, and the base labeling sites of oysters were obtained and defined as a class of base labeling sites. Individual gene samples of four other farmed oyster species were extracted from the farming logs and the same bioassay process was performed to obtain the base labeling sites of different farmed oyster parents, which were defined as Class II base labeling sites. Based on the hybridization expectation of oyster, an isotopic sliding window and sliding step size for base alignment were constructed. Within the isotopic sliding window, the variation difference of COI genetic marker sequences after alignment of Class I and Class II base labeling sites was analyzed. The Kimura 2-Parameter distance calculation framework was used for cumulative calculation to obtain the genetic substitution feature value of the window. Based on the sliding step size, the above steps of mutation analysis and genetic substitution feature calculation are repeated by moving the sliding window from the sequence start point backward to the same site until the entire COI sequence is covered. The genetic distance vector is obtained, and the COI genetic distance of the oyster sample relative to the cultured oyster sample is determined based on the genetic distance vector.

[0019] It should be noted that by performing multiple sequence mutation tests on the COI genetic markers of individual oysters and completing the labeling of base sites through explicit insertion operations, the inconsistency of sites caused by insertion, deletion or sequencing offset between different individuals is eliminated. This results in a set of COI base labeling sites with consistent structure and high comparability, ensuring that the genetic distance difference comes from real genetic differentiation and ensuring the accuracy and reproducibility of subsequent COI genetic distance calculations. The isolocate sliding window is used to capture the genetic differentiation intensity of different sequence segments of COI, accurately identifying conserved and hypervariable regions of locus function. The sliding step size is used to control the spatial sequence scale of COI genetic distance measurement, thus establishing a fine-grained balance between resolution and statistical stability. The overall genetic difference of the COI sequence is decomposed into the replacement feature values ​​of multiple local windows, shifting the genetic difference of COI from "overall average" to "local fine characterization". The flexible traversal sliding of the isolocate sliding window by the sliding step size effectively avoids the large-scale information redundancy or loss of details introduced by the traditional fixed window scale, thereby reflecting the contribution of transformation mutations and transversion mutations to different internal difference hotspots of genetic distance, improving the biological rationality of genetic distance assessment, making the measurement results closer to the real molecular evolution process, and providing a more refined genetic basis for the prediction of hybrid vigor.

[0020] Furthermore, based on the expected hybridization of the oyster *Crassostrea gigas*, an isotopic sliding window and sliding step size for base alignment are constructed. Within the isotopic sliding window, the variation differences of the COI genetic marker sequences after alignment of the first-class and second-class base labeling sites are analyzed, and the Kimura 2-Parameter distance calculation framework is used for cumulative calculation to obtain the genetic substitution feature value of the window. Specifically, this includes the following steps: Obtain hybridization expectation information of oysters, construct an isolocus sliding window for base reference alignment based on the genetic calculation granularity of COI genetic distance according to the hybridization expectation information, and preset the sliding step size according to the gene control resolution specified by the hybridization expectation information. A base sequence alignment space is constructed, and first-class base labeling sites and second-class base labeling sites are aligned in the base sequence alignment space. The Kimura 2-Parameter distance calculation architecture is introduced and coupled into the base sequence alignment space. Based on the sliding window of the equiloculation, the corresponding COI genetic marker sequence fragments between the aligned oyster samples and the cultured oyster samples are extracted from the window edge interval. A subset of COI alignment sequences under the sliding window of the equiloculation is established. Functional analysis and information entropy variation frequency weighting are performed on the base labeling sites of each window alignment interval in the subset of COI alignment sequences, and the mutation weight value of each base labeling site group is output. Based on the mutation weight values, the substitution differences of COI genetic marker sequences between the oyster sample and the cultured oyster sample are calculated by accumulating the values ​​within the current window using the Kimura 2-Parameter distance calculation architecture, and the genetic substitution feature value of the window is obtained.

[0021] It should be noted that the Kimura 2-Parameter (K2P) model is an evolutionary model of DNA sequence base substitutions, designed to describe the probability of base mutations occurring during the long evolutionary process of DNA sequences. It is one of the intuitive techniques for quantifying genetic distance. By aligning base sites one by one among different individuals, a unified site coordinate system, i.e., a base sequence alignment space, is established, making the weights of each site comparable. This method, by coupling the Kimura 2-Parameter distance calculation model within the base sequence alignment space, ensures that site mutations in all base sequences are compared under the same coordinate system and the same evolutionary model assumptions. This effectively distinguishes the different genetic evolutionary weights of transition mutations and transversion mutations, avoiding the large evolutionary bias caused by traditional simple base sequence difference counting. This makes the COI genetic distance calculation between the oyster *Crassostrea gigas* and different cultured oyster individuals more reliable for assessing hybridization affinity. For each base-labeled site, correlation analysis of functional attributes and variation frequency weighting based on information entropy assessment are performed to set genetic difference weights. Mutation weights highlight the differentiated contributions of different base mutations in genetic distance calculations, emphasizing high-variability, high-information, or functionally sensitive sites while effectively suppressing the influence of noise sites. This method overcomes the limitation of traditional genetic distance techniques that treat all sequence mutations equally, improving the ability to identify key genetic difference sites and enhancing the biological interpretability of hybridization potential assessment. Genetic substitution eigenvalues ​​reflect the true evolutionary and functional differences of local COI sequences, rather than simple average distance indicators. This method enables refined calculation of sequence site mutation substitutions in the COI genetic distance between *Crassostrea gigas* and different cultured oysters, providing high-resolution intraspecific genetic difference vectors for subsequent prediction of hybrid vigor in *Crassostrea gigas*, strengthening the reliability and discriminative ability of hybrid vigor determination.

[0022] Furthermore, S02 specifically includes the following steps: Obtain the breeding memo log of oyster hybridization, and extract historical hybridization survival data between oyster samples and cultured oyster samples within a preset time period, historical COI genetic variation data of each sample when different historical hybridization survival data are generated, and perturbation factor data captured during the hybridization process from the breeding memo log. Historical COI genetic variation data is defined as the independent variable and historical hybridization survival data is defined as the dependent variable. The marginal distribution between each independent variable and the dependent variable is constructed using the nonparametric distribution theorem to generate a marginal-dependent distribution model. The perturbation factor data is defined as a covariate. The probability of each historical hybridization survival data being treated by the covariate under the premise of dependence on different historical COI genetic variation data is estimated by the marginal-dependent distribution model, and the propensity of the regulatory residual of the perturbation factor data is obtained. Based on the probability of the controlled residual tendency, the perturbation factor data is introduced into the crossover distribution difference for each historical crossover survival data to calculate the tendency, and the perturbation tendency score is output. Based on the perturbation tendency score, all historical COI genetic variation data and historical hybrid survival data are jointly mapped to the [0,1] observation space for probability integral transformation, forming pseudo-observation samples of COI genetic variation-oyster hybrid survival advantage and corresponding observation dependence features. An error regression framework is introduced, and a conditional dependency function is preset based on the observation dependency characteristics. Based on the conditional dependency function, potential hybrid vigor is inferred by combining pseudo-observation samples and historical hybridization survival data in the error regression framework, thus generating the COI genetic-hybrid vigor regression model.

[0023] It should be noted that the hybridization breeding process of Crassula spinosa is generally subject to multi-dimensional perturbations such as hybridization environment, stocking density, and hybridization techniques. Traditional regression prediction models, when exploring the genetic variation of oyster COI, fail to incorporate the residual jitter of hybridization effects introduced by these perturbations. This leads to a significant cumulative cascading distribution bias in the regression training of hybridization survival rate data, resulting in severely insufficient predictive ability and a lack of accuracy for the hybrid vigor regression model. To address this, our method first treats historical COI genetic variation data as the independent variable and historical hybridization survival data as the dependent variable for training input. The historical COI genetic variation data includes variant sequence data, variant identification data, annotation description data, derivation analysis data, and statistical data; the historical hybridization survival data includes survival status data, survival time data, individual characteristic change data, and mortality index data. Subsequently, the marginal distribution patterns of the dependent variable and each independent variable are characterized by nonparametric distribution theorems, thereby capturing the independent empirical distribution estimate between COI genetic variation and hybridization survival without presupposing the specific correlation distribution forms between variables. This describes the nonlinear and asymmetric hybridization causal dependence structure, avoiding the regression guidance and limitations of traditional linear or normal assumptions on biological genetic data. The nonparametric distribution theorem adopts empirical distribution. Next, the perturbation factor data during the hybridization process is introduced as a covariate into the marginal-dependent distribution model to estimate the probability that each historical hybridization survival result is affected by the perturbation factor under different COI genetic variation dependence conditions. This maps multidimensional covariate information into a single probability index, eliminates non-random selection bias, and quantifies the intervention and control strength of non-genetic factors in manipulating the independent variable toward the hybridization result of the dependent variable. Based on this tendency probability, the hybridization distribution difference tendency introduced by the covariate is further calculated. The perturbation tendency score visualizes the distribution shift interference of the perturbation factor on the hybridization survival result, thereby assigning a corresponding regression treatment tendency level to each hybridization survival data sample, automatically weakening the interference of strong perturbation samples, achieving effective decoupling of genetic effect contribution and environmental or technical operation interference, preventing the misjudgment of hybridization survival due to exogenous perturbation as genetic advantage for superposition, and improving the purity and accuracy of inference of hybridization advantage in Crassula spinosa.

[0024] It should be noted that the original data samples (historical COI genetic variation data and historical hybrid survival data) are mapped to a unified observation space of [0,1] based on the perturbation tendency score. This eliminates differences in dimensions and scales, making data from different sources and with different distribution characteristics comparable in the same probability space. This forms standardized pseudo-observation samples and observation dependency features, ensuring that subsequent conditional dependency functions only characterize the dependency structure of distribution shifts, unaffected by marginal distributions. This lays a standardized foundation for regression inference based on COI genetic distance and improves the generalization ability of the regression model. This method automatically weighs the tendency of perturbation factors to intervene in the residuals of hybrid survival sample regression during the regression training of historical hybrid data, considering the dependency structure between independent and dependent variable data. This allows the regression inference training of dependent and independent variables to fully consider the survival interference contribution brought by external perturbations. Compared with traditional regression model training methods, this significantly improves the accuracy and reliability of hybrid vigor prediction based on COI genetic distance.

[0025] Furthermore, S03 specifically includes the following steps: By obtaining the threshold constraint range of COI genetic distance driving the survival rate improvement of the hybrid group of oyster, the COI genetic-hybrid vigor regression model is switched in different ranges according to the boundary of the threshold constraint range, and a multidimensional threshold structure driven by COI genetic distance is constructed. Based on the multidimensional threshold structure, the COI genetic distance is imported into the COI genetic-hybrid vigor regression model for regression prediction analysis, outputting a series of hybrid vigor regression coefficients. At the same time, the quantiles of the COI genetic distance are calculated, and the hybrid vigor regression coefficients are sorted in descending order according to the quantiles to generate a sparse regression path for hybrid vigor. After sorting in descending order, extreme coefficient samples on the sparse regression path are removed, and one or more regression coefficients of non-extreme samples are obtained, which are defined as candidate sparse regression coefficients. When the COI genetic distance is within the threshold constraint range, it is assumed that the survival rate of the hybrid group of the oyster sample increases by a positive percentage compared to the cultured oyster sample; when the COI genetic distance is not within the threshold constraint range, it is considered that the survival rate of the hybrid group does not change or decreases by a negative percentage, forming the threshold regression logic criterion for hybrid vigor. A log-likelihood algorithm is introduced. Based on the threshold regression logic criterion, the regression coefficient of each candidate coefficient is sample-likelihood fitting in the log-likelihood algorithm to generate the hybrid vigor log-likelihood value of each candidate sparse regression coefficient based on threshold logic control. The regression trend of each hybrid vigor log-likelihood value is jointly predicted according to the COI genetic distance to obtain the hybrid vigor rate of oyster.

[0026] It should be noted that numerous studies have shown that when the Kimura 2-Parameter distance between the oyster and the farmed oyster is within the threshold range of 0.18–0.24, the survival rate of the hybrid group can be increased by 18%. However, traditional regression prediction methods often lack the ability to explore hybrid survival rates within specific threshold structures, making it difficult to accurately measure the granularity of hybrid vigor at different COI genetic distances. This leads to significant regression decision errors for hybrid vigor corresponding to COI genetic distances within different threshold ranges. To address this, this method switches the operating mode of the COI genetic-hybrid vigor regression model within different genetic distance ranges based on the upper and lower boundaries of the threshold range examples mentioned above. This is used to hypothesize that a certain genetic distance variable will trigger a structural change in the intrinsic regression relationship after reaching a specific threshold level. This constructs a multidimensional threshold structure dominated by COI genetic distance as a precursor. This multidimensional threshold structure can explicitly constrain the "implicit regression inflection point" in the COI genetic-hybrid vigor regression model, indirectly restricting the interval sensitivity of genetic distance to hybrid vigor, enabling the regression model to exhibit a differentiated response mechanism for different genetic distance ranges. The quantiles of the COI genetic distance are used to identify and rank a reasonable threshold search space, preserving the representative and effective sparse components of the regression prediction, i.e., the sparse regression path. This sparse regression path reveals the combination of the distribution characteristics of genetic distance and the strength of regression coefficients, further highlighting key genetic intervals and parameters that significantly contribute to hybrid vigor. This method then removes extreme regression coefficient samples, retaining one or more non-extreme regression coefficients as candidate sparse regression coefficients. This avoids imbalanced regression coefficient samples caused by excessively extreme thresholds, thereby eliminating the instability of regression predictions caused by sparse sample noise or extreme genetic distances. Furthermore, by planning the threshold regression logic criteria for hybrid vigor, the continuous regression problem is transformed into an interval response mechanism constrained by logical rules. This prevents the phenomenon of incorrectly amplifying hybrid vigor due to the irregular and free interpretation of COI genetic distance in traditional regression prediction methods, improving the accuracy of subsequent vigor inference based on threshold structure constraints and fluctuations in hybrid survival performance. This method embeds a specific threshold interval structure into the COI genetic-hybrid vigor regression model, thereby accurately controlling the assignment of hybrid vigor according to the regression decision logic of survival performance scale under the constraint of the threshold structure when predicting COI genetic distance. This improves the accuracy and reliability of hybrid vigor prediction and provides a scientific evaluation index reference for hybrid breeding of Crassula spinosa.

[0027] Furthermore, S04 specifically includes the following steps: If the hybrid vigor rate is less than the preset hybrid vigor rate, the hybridization selection of the cultured oyster sample is ignored; if the hybrid vigor rate is greater than the preset hybrid vigor rate, several candidate parents to be hybridized and their corresponding actual phenotypic data are obtained for the cultured oyster sample, and the COI genetic distance matrix between the candidate parents to be hybridized and the oyster sample is constructed based on the COI genetic distance. Based on breeding experience cases, query actual phenotypic data to obtain prior distribution information on phenotypic transfer between candidate parents and different oysters. Based on the prior distribution information on phenotypic transfer, construct intraspecific gene clusters of parents. Combine the COI genetic distance matrix and threshold constraint interval to assign each parent to intraspecific gene clusters and change the corresponding fertile cluster center to generate a hybridization fertility index. Based on the hybridization fertility index, select acceptable intraspecific fixed parents to be hybridized. The current growth trait index of the fixed parent in the species to be hybridized is obtained and defined as the first growth trait index. Based on the breeding experience case search, the maximum bottleneck resistance coefficient that restricts the complementary expression of growth traits of the fixed parent in the species to be hybridized for the oyster sample is determined at the gene level. Based on the maximum bottleneck resistance coefficient, the complementary bottleneck function of the parents is constructed, and the current growth trait index of the oyster sample is obtained and defined as the second growth trait index. The complementary dominance of each first growth trait index to the hybridization genetic expression between the second growth trait index is evaluated to obtain the trait complementary coupling value. If the complementary coupling value of a trait is lower than a preset threshold, a penalty is applied to the first growth trait index corresponding to the complementary coupling value of the trait, a penalty coupling criterion term is constructed, and the penalty coupling criterion term is introduced into the complementary bottleneck function. A genetic algorithm is introduced to construct a coupled solver. The coupled solver is used to iteratively update the solutions for each first growth trait index in order to identify the optimality of the coupling for the complementarity of the spiny oyster sample after penalty coordination in the complementary bottleneck function corresponding to the fixed parents in each hybrid species, and to generate the complementarity strength of traits. Only the cultured oyster individuals corresponding to the fixed parent species of the species to be hybridized with the maximum complementary strength of the trait are extracted and paired with the spiny oyster individuals to output the hybridization combination configuration scheme of spiny oyster.

[0028] It should be noted that when the hybrid vigor rate is less than the preset hybrid vigor rate, it indicates that the hybrid survival rate between the *Crassostrea gigas* sample and the cultured oyster sample of that species is not up to standard, and the hybrid performance is low. In this case, the cultured oyster sample of that species will not be selected as the hybrid parent. When the hybrid vigor rate is greater than the preset hybrid vigor rate, it indicates that the hybrid survival rate between the *Crassostrea gigas* sample and the cultured oyster sample of that species is up to standard, and it shows good hybrid performance. Therefore, it will be selected. However, some of the parents have excessively large or small COI genetic distances from the *Crassostrea gigas*, resulting in extremely small differences in the appropriate level of impurity vigor, leading to poor hybrid fertility and making it difficult to meet the effect combination of excellent hybrid performance of *Crassostrea gigas*. To address this, this method constructs intraspecific gene clusters of cultured oyster parents based on phenotypic transfer priors. By combining the COI genetic distance matrix with threshold constraint intervals, the parents are dynamically allocated to each gene cluster, and the fertile cluster centers are updated. This allows for the assessment of hybridization fertility between Crassula spinosa and these selected parents. Thus, under the dual constraints of COI genetic distance and phenotypic compatibility, intraspecific fixed parents with acceptable hybridization and stable genetic structure are preliminarily screened. This avoids hybridization failures caused by unclear advantages of genetic distance and reproductive incompatibility, thereby improving the hybridization success rate and breeding stability of Crassula spinosa.

[0029] It should be noted that after selecting an acceptable range for intraspecific hybridization (fixed parents within the species), these parents still exhibit minimal differences in growth traits with the Crassula spinosa. This means that the improvement and complementarity of growth traits in Crassula spinosa from cultured oysters is poor, failing to enhance the breeding performance of Crassula spinosa. Therefore, it is necessary to select cultured oyster individuals with strong complementarity to Crassula spinosa traits for hybridization. To address this, this method constructs a complementarity bottleneck function for the parents based on the maximum bottleneck resistance coefficient. The maximum bottleneck resistance coefficient quantifies the potential genetic limitations or expression limits of different parents during trait complementarity, preventing the ineffective enhancement of traits lacking complementarity potential. The complementarity bottleneck function represents the weakest link in the complementary ability of the fixed parents. Furthermore, by evaluating the hybridization complementarity coupling dominance of the current growth traits of each fixed parent to be hybridized to the genetic expression of growth traits in Crassula spinosa, i.e., the trait complementarity coupling value, this trait complementarity coupling value quantitatively characterizes the true complementary ability of cultured oyster parent traits to enhance Crassula spinosa traits. If the trait complementarity coupling value is lower than a preset threshold, it indicates that the complementarity reflected by this growth trait index is poor. This suggests that the fixed parents within the species to be hybridized have a certain proportion of weak trait complementarity determining components. Therefore, additional penalties are applied to these components to suppress the weighting of parental traits that contribute insufficiently to the hybridization trait or may have a negative effect during the optimization process. This effectively avoids the problem of local trait advantages masking overall insufficient complementarity. This method, under the premise of clearly defining the COI genetic distance and the fertile hybridization range, can screen out the cultured parent individuals with the strongest trait complementarity for hybridization with *Crassostrea gigas*, thereby improving the overall performance of the final hybrid combination, ensuring the hybridization quality of *Crassostrea gigas*, and achieving a scientific decision-making effect in breeding configuration.

[0030] Furthermore, the method involves obtaining prior phenotypic distribution information of the candidate parents for hybridization and different oyster species by querying actual phenotypic data based on breeding experience cases, constructing intraspecific gene clusters of the parents based on the prior phenotypic distribution information, assigning each parent to an intraspecific gene cluster by combining the COI genetic distance matrix and threshold constraint intervals, and changing the corresponding fertile cluster centers to generate a hybridization fertility index. Acceptable intraspecific fixed parents for hybridization are then selected based on the hybridization fertility index. This process specifically includes the following steps: Based on big data, we can obtain breeding experience cases of candidate parents for hybridization, and obtain the prior distribution information of phenotypic transfer of candidate parents for hybridization with different oyster varieties by querying the actual phenotypic data of the breeding experience cases. Based on the prior distribution information of phenotypic transfer, intraspecific gene clusters and corresponding fertile cluster centers of different candidates for hybridization are constructed. The breeding advantage distance quantification of the COI genetic distance matrix and the threshold constraint interval is combined to appropriately allocate each candidate parent to each intraspecific gene cluster. During the allocation process, the fertile cluster centers of all parents assigned to the gene clusters of that species are replaced, and a weighted average is performed according to the normalized weights to output the current scaling stability of the fertile cluster centers. Repeat the above steps to iterate and calculate various intragene clusters. If the current scaling stability is greater than 0, stop the iteration operation immediately and finally output the squared error of the intragene cluster to which the oyster sample belongs. Based on the squared error, the hybridization combination between the oyster sample and each candidate parent to be hybridized is quantitatively evaluated based on the appropriate hybridization phenotypic performance at the advantage threshold, and a hybridization fertility index is generated. If the hybridization fertility index is greater than the maximum hybridization fertility index, the candidate parent to be hybridized corresponding to the hybridization fertility index is extracted and marked to obtain one or more fixed parents within the species to be hybridized.

[0031] It should be noted that by obtaining the prior distribution information of phenotypic transfer in hybridization between candidate parents and different Crassula spinosa varieties, this prior distribution information represents the probability distribution of known phenotypic feature transmission in hybridization between different cultured parents and Crassula spinosa. This prior distribution information serves as the long-term breeding experience base for hybridization combination decisions, providing a clustering link for the correct phenotypic vector in intraspecific hybridization selection for Crassula spinosa fertility. Intraspecific gene clusters represent the phenotypic semantic search space of each candidate parent's gene pool relative to Crassula spinosa genes, while the fertile cluster center anchors the initial intraspecific gene feasibility consideration allocation point within this semantic search space under the known premise of prior phenotypic feature transmission. By combining the COI genetic distance matrix with threshold constraint intervals to allocate parents to various intraspecific gene clusters, the transitional state of parent-species Crassostrea gigas gene combinations is captured. This ensures appropriate genetic distances while achieving reasonable aggregation of parents within the species and Crassostrea gigas, avoiding combinations with excessively small genetic distances, which may result in the cultured parents and Crassostrea gigas originating from the same lineage, leading to insignificant hybrid vigor. Conversely, selecting parental combinations with excessively large genetic distances may lead to reproductive incompatibility and phenotypic genetic abnormalities. The study comprehensively considers both the fertility of the genetic structure and phenotypic compatibility. The fertile cluster centers are iterated by adjusting their positions to adapt to the Crassostrea gigas sample structure, enabling the fertile cluster centers to reflect in real-time the relative fertility nodes of the overall genetic and phenotypic characteristics of the current intraspecific gene combinations of parents and Crassostrea gigas. Scaling stability measures the convergence stability of the cluster structure.

[0032] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for predicting and configuring hybrid vigor in *Crassostrea gigas* based on COI genetic distance, characterized in that, Includes the following steps: S01: Collect and detect the base labeling sites of oysters and various farmed oysters by biological sequence detection, construct an isotope sliding window for base alignment, and use the K2P distance calculation architecture in the isotope sliding window to calculate the variation difference after the alignment of two base labeling sites on the COI genetic sequence, so as to obtain the COI genetic distance of oyster samples compared with farmed oyster samples. S02: Obtain historical hybridization survival data, historical COI genetic variation data, and perturbation factor data of oyster samples and cultured oyster samples. Estimate the propensity of the regulatory residuals that are passively treated with covariates of the perturbation factor data. Based on the propensity of the regulatory residuals, establish an observation-dependent regression relationship between historical hybridization survival data and historical COI genetic variation data to generate a COI genetic-hybridization advantage regression model. S03: Based on the threshold constraint interval of COI genetic distance driving the survival rate improvement of the hybrid group of oyster, a multidimensional threshold structure is constructed. Based on the multidimensional threshold structure, the COI genetic distance is introduced into the COI genetic-hybrid vigor regression model to perform threshold logic regulation regression prediction analysis on the hybrid vigor of oyster, and the hybrid vigor rate of oyster is obtained. S04: Based on the hybrid vigor, select candidate parents of cultured oysters and construct a COI genetic distance matrix. Combine the COI genetic distance matrix with the intraspecific fertility analysis of the threshold constraint interval to screen out the fixed parents to be hybridized. By coupling the current growth trait indicators, infer the trait complementarity bottleneck of the fixed parents to be hybridized for the oyster. Select cultured oyster individuals with the greatest trait complementarity intensity for hybridization combination pairing.

2. The method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance according to claim 1, characterized in that, S01 specifically includes the following steps: Individual gene samples of oysters were collected, and multiple sequence mutation tests were performed on the COI genetic markers of the individual gene samples using a biodetection method. Explicit insertion and deletion sites were identified, and the base labeling sites of oysters were obtained and defined as a class of base labeling sites. Individual gene samples of four other farmed oyster species were extracted from the farming logs and the same bioassay process was performed to obtain the base labeling sites of different farmed oyster parents, which were defined as Class II base labeling sites. Based on the hybridization expectation of oyster, an isotopic sliding window and sliding step size for base alignment were constructed. Within the isotopic sliding window, the variation difference of COI genetic marker sequences after alignment of Class I and Class II base labeling sites was analyzed. The Kimura 2-Parameter distance calculation framework was used for cumulative calculation to obtain the genetic substitution feature value of the window. Based on the sliding step size, the above steps of mutation analysis and genetic substitution feature calculation are repeated by moving the sliding window from the sequence start point backward to the same site until the entire COI sequence is covered. The genetic distance vector is obtained, and the COI genetic distance of the oyster sample relative to the cultured oyster sample is determined based on the genetic distance vector.

3. The method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance according to claim 2, characterized in that, The method involves constructing an isolocution sliding window and sliding step size for base alignment based on the expected hybridization of the oyster *Crassostrea gigas*. Within the isolocution sliding window, the variation differences of the COI genetic marker sequences after alignment of Class I and Class II base marker sites are analyzed. The genetic substitution feature values ​​of the window are obtained using the Kimura 2-Parameter distance calculation framework for cumulative calculation. The specific steps include: Obtain hybridization expectation information of oysters, construct an isolocus sliding window for base reference alignment based on the genetic calculation granularity of COI genetic distance according to the hybridization expectation information, and preset the sliding step size according to the gene control resolution specified by the hybridization expectation information. A base sequence alignment space is constructed, and first-class base labeling sites and second-class base labeling sites are aligned in the base sequence alignment space. The Kimura 2-Parameter distance calculation architecture is introduced and coupled into the base sequence alignment space. Based on the sliding window of the equiloculation, the corresponding COI genetic marker sequence fragments between the aligned oyster samples and the cultured oyster samples are extracted from the window edge interval. A subset of COI alignment sequences under the sliding window of the equiloculation is established. Functional analysis and information entropy variation frequency weighting are performed on the base labeling sites of each window alignment interval in the subset of COI alignment sequences, and the mutation weight value of each base labeling site group is output. Based on the mutation weight values, the substitution differences of COI genetic marker sequences between the oyster sample and the cultured oyster sample are calculated by accumulating the values ​​within the current window using the Kimura 2-Parameter distance calculation architecture, and the genetic substitution feature value of the window is obtained.

4. The method for predicting and combining hybrid vigor in *Crassostrea gigas* based on COI genetic distance according to claim 1, characterized in that, S02 specifically includes the following steps: Obtain the breeding memo log of oyster hybridization, and extract historical hybridization survival data between oyster samples and cultured oyster samples within a preset time period, historical COI genetic variation data of each sample when different historical hybridization survival data are generated, and perturbation factor data captured during the hybridization process from the breeding memo log. Historical COI genetic variation data is defined as the independent variable and historical hybridization survival data is defined as the dependent variable. The marginal distribution between each independent variable and the dependent variable is constructed using the nonparametric distribution theorem to generate a marginal-dependent distribution model. The perturbation factor data is defined as a covariate. The probability of each historical hybridization survival data being treated by the covariate under the premise of dependence on different historical COI genetic variation data is estimated by the marginal-dependent distribution model, and the propensity of the regulatory residual of the perturbation factor data is obtained. Based on the probability of the controlled residual tendency, the perturbation factor data is introduced into the crossover distribution difference for each historical crossover survival data to calculate the tendency, and the perturbation tendency score is output. Based on the perturbation tendency score, all historical COI genetic variation data and historical hybrid survival data are jointly mapped to the [0,1] observation space for probability integral transformation, forming pseudo-observation samples of COI genetic variation-oyster hybrid survival advantage and corresponding observation dependence features. An error regression framework is introduced, and a conditional dependency function is preset based on the observation dependency characteristics. Based on the conditional dependency function, potential hybrid vigor is inferred by combining pseudo-observation samples and historical hybridization survival data in the error regression framework, thus generating the COI genetic-hybrid vigor regression model.

5. The method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance according to claim 1, characterized in that, S03 specifically includes the following steps: By obtaining the threshold constraint range of COI genetic distance driving the survival rate improvement of the hybrid group of oyster, the COI genetic-hybrid vigor regression model is switched in different ranges according to the boundary of the threshold constraint range, and a multidimensional threshold structure driven by COI genetic distance is constructed. Based on the multidimensional threshold structure, the COI genetic distance is imported into the COI genetic-hybrid vigor regression model for regression prediction analysis, outputting a series of hybrid vigor regression coefficients. At the same time, the quantiles of the COI genetic distance are calculated, and the hybrid vigor regression coefficients are sorted in descending order according to the quantiles to generate a sparse regression path for hybrid vigor. After sorting in descending order, extreme coefficient samples on the sparse regression path are removed, and one or more regression coefficients of non-extreme samples are obtained, which are defined as candidate sparse regression coefficients. When the COI genetic distance is within the threshold constraint range, it is assumed that the survival rate of the hybrid group of the oyster sample increases by a positive percentage compared to the cultured oyster sample; when the COI genetic distance is not within the threshold constraint range, it is considered that the survival rate of the hybrid group does not change or decreases by a negative percentage, forming the threshold regression logic criterion for hybrid vigor. A log-likelihood algorithm is introduced. Based on the threshold regression logic criterion, the regression coefficient of each candidate coefficient is sample-likelihood fitting in the log-likelihood algorithm to generate the hybrid vigor log-likelihood value of each candidate sparse regression coefficient based on threshold logic control. The regression trend of each hybrid vigor log-likelihood value is jointly predicted according to the COI genetic distance to obtain the hybrid vigor rate of oyster.

6. The method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance according to claim 1, characterized in that, S04 specifically includes the following steps: If the hybrid vigor rate is less than the preset hybrid vigor rate, the hybridization selection of the cultured oyster sample is ignored; if the hybrid vigor rate is greater than the preset hybrid vigor rate, several candidate parents to be hybridized and their corresponding actual phenotypic data are obtained for the cultured oyster sample, and the COI genetic distance matrix between the candidate parents to be hybridized and the oyster sample is constructed based on the COI genetic distance. Based on breeding experience cases, query actual phenotypic data to obtain prior distribution information on phenotypic transfer between candidate parents and different oysters. Based on the prior distribution information on phenotypic transfer, construct intraspecific gene clusters of parents. Combine the COI genetic distance matrix and threshold constraint interval to assign each parent to intraspecific gene clusters and change the corresponding fertile cluster center to generate a hybridization fertility index. Based on the hybridization fertility index, select acceptable intraspecific fixed parents to be hybridized. The current growth trait index of the fixed parent in the species to be hybridized is obtained and defined as the first growth trait index. Based on the breeding experience case search, the maximum bottleneck resistance coefficient that restricts the complementary expression of growth traits of the fixed parent in the species to be hybridized for the oyster sample is determined at the gene level. Based on the maximum bottleneck resistance coefficient, the complementary bottleneck function of the parents is constructed, and the current growth trait index of the oyster sample is obtained and defined as the second growth trait index. The complementary dominance of each first growth trait index to the hybridization genetic expression between the second growth trait index is evaluated to obtain the trait complementary coupling value. If the complementary coupling value of a trait is lower than a preset threshold, a penalty is applied to the first growth trait index corresponding to the complementary coupling value of the trait, a penalty coupling criterion term is constructed, and the penalty coupling criterion term is introduced into the complementary bottleneck function. A genetic algorithm is introduced to construct a coupled solver. The coupled solver is used to iteratively update the solutions for each first growth trait index in order to identify the optimality of the coupling for the complementarity of the spiny oyster sample after penalty coordination in the complementary bottleneck function corresponding to the fixed parents in each hybrid species, and to generate the complementarity strength of traits. Only the cultured oyster individuals corresponding to the fixed parent species of the species to be hybridized with the maximum complementary strength of the trait are extracted and paired with the spiny oyster individuals to output the hybridization combination configuration scheme of spiny oyster.

7. The method for predicting and combining hybrid vigor in Crassula spinosa based on COI genetic distance according to claim 6, characterized in that, The process involves obtaining prior phenotypic distribution information on candidate parents for hybridization and different oyster species by querying actual phenotypic data based on breeding experience cases. Intraspecific gene clusters of the parents are constructed based on this prior phenotypic distribution information. Each parent is then assigned to an intraspecific gene cluster using the COI genetic distance matrix and threshold constraint intervals, and the corresponding fertile cluster centers are updated to generate a hybridization fertility index. Acceptable intraspecific fixed parents for hybridization are then selected based on this hybridization fertility index. The specific steps include: Based on big data, we can obtain breeding experience cases of candidate parents for hybridization, and obtain the prior distribution information of phenotypic transfer of candidate parents for hybridization with different oyster varieties by querying the actual phenotypic data of the breeding experience cases. Based on the prior distribution information of phenotypic transfer, intraspecific gene clusters and corresponding fertile cluster centers of different candidates for hybridization are constructed. The breeding advantage distance quantification of the COI genetic distance matrix and the threshold constraint interval is combined to appropriately allocate each candidate parent to each intraspecific gene cluster. During the allocation process, the fertile cluster centers of all parents assigned to the gene clusters of that species are replaced, and a weighted average is performed according to the normalized weights to output the current scaling stability of the fertile cluster centers. Repeat the above steps to iterate and calculate various intragene clusters. If the current scaling stability is greater than 0, stop the iteration operation immediately and finally output the squared error of the intragene cluster to which the oyster sample belongs. Based on the squared error, the hybridization combination between the oyster sample and each candidate parent to be hybridized is quantitatively evaluated based on the appropriate hybridization phenotypic performance at the advantage threshold, and a hybridization fertility index is generated. If the hybridization fertility index is greater than the maximum hybridization fertility index, the candidate parent to be hybridized corresponding to the hybridization fertility index is extracted and marked to obtain one or more fixed parents within the species to be hybridized.

Images