Molecular marker assisted high yield strain directional breeding of dictyophora rubrovolvata

Abstract

The application discloses a molecular marker assisted high-yield Phyllostachys annenii strain directional breeding and screening method, and relates to the technical field of molecular marker assisted breeding.The method comprises the following steps: obtaining amplification band data of each test strain at multiple molecular marker sites and agronomic trait data of each test strain; performing point-by-point difference comparison on the amplification band data and the starting strain to construct a mutagenesis difference fingerprint vector; performing molecular marker detection on multiple single colonies of the same test strain, calculating a consistency index and correcting genotype data; performing point-by-point correlation analysis on the corrected genotype data as the independent variable and the agronomic trait data as the dependent variable, screening a core molecular marker combination and determining a contribution weight; and calculating a breeding value score according to the genotype at the core marker site and the contribution weight to screen a candidate strain.The application solves the technical problems of low screening efficiency of spaceflight mutagenesis Phyllostachys annenii mutants and that conventional molecular marker methods do not consider the particularity of mutagenesis materials.

Description

Technical Field

[0001] This invention relates to the field of molecular marker-assisted breeding technology, specifically to a method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* var. *red* assisted by molecular markers. Background Technology

[0002] Red-topped bamboo fungus is a rare edible fungus with high economic and medicinal value. Space mutagenesis breeding involves sending fungal strains into space aboard a spacecraft, where high-energy rays, microgravity, and high vacuum induce gene mutations and chromosomal aberrations in the biological material. After returning to Earth, mutants with superior traits are selected. The mutation types and sites generated by space mutagenesis are random and diverse, enabling the acquisition of mutant traits that are difficult to obtain using conventional breeding methods.

[0003] Conventional marker-assisted breeding methods typically compare the differences in amplified bands between the tested strains and the starting strain to screen for molecular marker loci associated with the target trait for selection. However, the mutation stability of space-induced mutagenesis materials differs from that of conventional hybrid offspring. Mutants may exhibit instability due to somatic variation during subculture, leading to inconsistent genotype detection. Conventional methods do not perform consistency testing and correction on multiple single colonies within the same tested strain, potentially introducing genotype data errors. Furthermore, high yield in *Dictyophora indica* is a complex trait, with multiple traits such as mycelial growth rate, biological efficiency, fruiting body size, and stress resistance contributing differently to high yield. Conventional methods treat all traits as equally important for correlation analysis, failing to distinguish the varying contributions of different traits to the high-yield target, resulting in a low match between the screened marker loci and the actual breeding objective. Summary of the Invention

[0004] This invention provides a molecular marker-assisted method for targeted breeding of high-yielding strains of *Dictyophora indicum*, which addresses the technical problems of low screening efficiency of space-induced mutants of *Dictyophora indicum* and the failure of conventional molecular marker methods to consider the specific characteristics of the mutagenic materials.

[0005] This invention provides a molecular marker-assisted method for targeted breeding of high-yielding strains of *Dictyophora indicum*, the method comprising:

[0006] Acquire amplification band data of each tested strain at multiple molecular marker sites, as well as agronomic trait data of each tested strain;

[0007] The amplification band data of each tested strain were compared with the amplification band data of the starting strain at each site to construct the mutagenic differential fingerprint vector of each tested strain.

[0008] Molecular marker detection was performed on multiple single colonies of the same test strain, and the consistency index of the detection results at each site was calculated. The genotype data of the test strain at the corresponding site was corrected based on the consistency index.

[0009] Using the corrected genotype data of each tested strain at each locus as independent variables and the agronomic trait data as dependent variables, a locus-by-locus association analysis was performed to screen out the core molecular marker combinations associated with the target agronomic traits and determine the contribution weight of each core marker locus.

[0010] Based on the genotype and corresponding contribution weight of each tested strain at the core marker site, the breeding value score of each tested strain is calculated, and candidate strains are screened based on the breeding value score.

[0011] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0012] First, this invention corrects the genotype data of the test strain by performing molecular marker detection on multiple single colonies of the same test strain and calculating the consistency index of each site. Sites with consistency below the threshold are excluded from subsequent analysis. Compared with conventional methods that directly use the results of a single test as genotype data, this invention can eliminate the error introduced by the inconsistency of genotype detection caused by the instability of somatic cell variation after space mutagenesis, and improve the reliability of genotype data.

[0013] Second, this invention determines the weighting coefficients based on the contribution of each agronomic trait to the high-yield breeding target, and uses the weighted agronomic trait data as the dependent variable in the association analysis. Compared with conventional methods that treat each trait as equally important, this invention makes the results of the association analysis more accurately reflect the actual contribution of each molecular marker site to the overall high-yield target, and improves the matching degree between the screened core molecular marker combination and the breeding target.

[0014] Third, this invention constructs a mutagenic differential fingerprint vector by comparing the site-by-site differences with the starting strain, which intuitively records the mutation status of each tested strain relative to the starting strain at each molecular marker site, providing molecular-level quantitative features covering three states: new bands, missing bands, and no mutations, for association analysis and breeding value scoring. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic flowchart of the molecular marker-assisted targeted breeding method for high-yielding strains of *Dictyophora indicum* provided in this embodiment of the invention.

[0017] Figure 2 This is a schematic diagram of the scheme for the molecular marker-assisted targeted breeding and selection method of high-yielding strains of *Dictyophora indicum* provided in the embodiments of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0019] Example 1, as Figure 1 , Figure 2 As shown, this invention provides a molecular marker-assisted method for targeted breeding of high-yielding strains of *Dictyophora indicum*, the method comprising:

[0020] S10: Obtain amplification band data of each tested strain at multiple molecular marker sites, as well as agronomic trait data of each tested strain;

[0021] The amplified band data from molecular marker sites are the fundamental molecular data for subsequent construction of mutagenic differential fingerprint vectors, correction of genotype data, and site-by-site association analysis. This step uses ISSR molecular marker technology to detect polymorphism in each tested strain. Primers were designed using conserved regions between repetitive sequences for PCR amplification. The presence or absence of amplified bands at each site was determined by gel electrophoresis to obtain 0 / 1 genotype data. Simultaneously, measured values ​​of agronomic traits such as mycelial growth rate, biological efficiency, fruiting body size, and stress resistance for each tested strain under standard cultivation conditions were also obtained. These agronomic trait data serve as target variables for association analysis, collectively constituting the phenotypic basis for subsequent breeding value scoring.

[0022] The specific implementation method is as follows:

[0023] Mycelial samples of each tested strain were obtained after space-induced mutagenesis. The tested strains were obtained by isolation and purification of *Dictyophora indica* mycelia induced by space mutagenesis and preliminary screening, including mutant strains and control strains without significant mutations. The starting strains were the original strains preserved before spaceflight.

[0024] Genomic DNA was extracted from each tested strain using the CTAB method. Mycelia of the tested strains were ground into powder in liquid nitrogen, and preheated CTAB extraction buffer was added to lyse the cells. Proteins were removed by chloroform extraction, and DNA was precipitated with isopropanol. After washing with ethanol, the DNA was dissolved in TE buffer. The concentration and purity of the DNA were measured using a UV spectrophotometer, and the concentration was uniformly diluted to 50 ng / μL for PCR amplification.

[0025] PCR amplification of DNA from each tested strain was performed using a pre-defined set of ISSR primers. ISSR primers are oligonucleotide sequences with 1 to 4 degenerate bases anchored at the 5' or 3' end of simple repeat sequences. The primer sequences were selected from the ISSR primer sequence set published by Columbia University, with 10 primers exhibiting high polymorphism chosen for amplification. The annealing temperature for each primer was set according to its GC content and length. Table 1, exemplarily, shows the 10 ISSR primer sequences and annealing temperatures selected in this embodiment of the invention.

[0026] Table 1: ISSR Primer Sequences and Annealing Temperatures

[0027] Primer number Primer sequence (5'→3') Annealing temperature (°C) UBC808 AGAGAGAGAGAGAGAGC 52 UBC811 GAGAGAGAGAGAGAGAC 52 UBC834 AGAGAGAGAGAGAGAGAGYT 50 UBC836 AGAGAGAGAGAGAGAGAGYA 51 UBC840 GAGAGAGAGAGAGAGAYT 45 UBC841 GAGAGAGAGAGAGAGAYC 47 UBC842 GAGAGAGAGAGAGAGAYG 46 UBC880 GGAGAGGAGAGGAGA 43 UBC888 BDBCACACACACACACA 45 UBC889 DBDACACACACACACAC 46

[0028] The PCR amplification reaction system included: 1 μL template DNA, 5 μL primers (10 μM), 12.5 μL 2×Taq PCR Master Mix, and 6.5 μL sterile deionized water. The amplification program was as follows: pre-denaturation at 94℃ for 5 minutes, followed by 35 cycles (denaturation at 94℃ for 30 seconds, annealing at the corresponding primer temperature for 60 seconds, extension at 72℃ for 45 seconds), and a final extension at 72℃ for 10 minutes, followed by storage at 4℃.

[0029] The amplified products were separated by 1.5% agarose gel electrophoresis, stained with gel red nucleic acid dye, and then imaged under ultraviolet light using a gel imaging system. For each tested strain, the electrophoretic pattern under each primer was analyzed lane by lane to determine the presence or absence of bands at each molecular weight position. A clearly visible band at a position with the same migration rate was recorded as 1, and no band was recorded as 0. The amplified sites of all primers were uniformly numbered according to primer number and site order to construct an amplification band matrix for each tested strain at all molecular marker sites. For example, 10 primers amplified a total of 128 polymorphic sites, and the amplification band matrix of the 36 tested strains was 36 rows × 128 columns.

[0030] Agronomic trait data were obtained for each tested strain. Specifically, agronomic trait data included mycelial growth rate, biological efficiency, fruiting body size, and stress resistance.

[0031] Mycelial growth rate was assessed by inoculating onto PDA plates and incubating in the dark at 25°C. Colony diameter was measured every 3 days using the streak method, and the increase in diameter during the linear growth phase was calculated by dividing by time, expressed in millimeters per day. Biological efficiency was measured by dry weight, calculated by weighing the fresh weight of the fruiting bodies, multiplying by the measured moisture content coefficient to obtain the dry weight, and dividing by the dry weight of the substrate, expressed as a percentage. Fruiting body size was comprehensively reflected by four indicators: stipe length, stipe diameter, veil length, and veil thickness. Stipe length was measured by randomly selecting at least 10 normal mature fruiting bodies from each tested strain, using calipers to measure the straight-line distance from the base of the stipe to the veil attachment point, and the arithmetic mean was taken, expressed in millimeters. Stipe diameter was measured by selecting the above fruiting bodies, using calipers to measure the cross-sectional diameter of the stipe at the middle, and the arithmetic mean was taken, expressed in millimeters. Veil length was measured by selecting the above fruiting bodies, using calipers to measure the straight-line distance from the veil attachment point to the edge of the veil, and the arithmetic mean was taken, expressed in millimeters. The thickness of the mycelial skirt was measured using calipers at the center of the selected fruiting bodies, and the arithmetic mean was taken, in millimeters. Stress resistance was quantified by the relative growth rate of mycelia under high-temperature stress. Inoculated PDA plates were placed in normal culture temperature (25℃) and under preset high-temperature stress conditions, and the linear growth rate of mycelia was measured at both temperatures. The preset high-temperature stress conditions were set based on the optimal temperature range and heat stress response characteristics of *Dictyophora indicum* mycelia. The optimal growth temperature range for *Dictyophora indicum* mycelia is 22℃ to 26℃. When the ambient temperature exceeds 30℃, mycelial growth begins to be significantly inhibited, and above 35℃, most strains stop growing or even die. For example, a culture temperature of 30℃ was used as the high-temperature stress condition. Stress resistance = (growth rate under high-temperature stress ÷ growth rate at normal temperature) × 100%. Higher stress resistance indicates less inhibition of growth under high-temperature stress and stronger adaptability to the cultivation environment.

[0032] The above four traits were measured for each tested strain in at least three independent cultivation batches, and the arithmetic mean of the measured values ​​from each batch was taken as the agronomic trait data of the tested strain. For example, a certain tested strain had a mycelial growth rate of 1.5 mm / d, a biological efficiency of 12.5%, a stipe length of 128 mm, a stipe diameter of 18.5 mm, a veil length of 85 mm, a veil thickness of 2.8 mm, and a stress resistance of 85%.

[0033] The following technical effects were achieved through this step:

[0034] ISSR molecular marker amplification and gel electrophoresis interpretation yielded a genotype matrix of 0 / 1 types at all molecular marker sites for each tested strain. This provided molecular-level input data for constructing a mutagenic differential fingerprint vector by comparing site-by-site differences with the starting strain. Standardized cultivation experiments were conducted to determine agronomic traits such as mycelial growth rate, biological efficiency, fruiting body size, and stress resistance, providing phenotypic target variables for association analysis and breeding value scoring.

[0035] S20: Compare the amplification band data of each tested strain with the amplification band data of the starting strain at each site to construct the mutagenic differential fingerprint vector of each tested strain.

[0036] Step S10 has obtained amplification band data for each tested strain and the originating strain. By comparing each tested strain with the originating strain site by site, the mutation status of each site can be classified as a new band, a missing band, or no mutation, and a mutagenic differential fingerprint vector can be constructed to intuitively record the genetic variation of each tested strain relative to the originating strain.

[0037] Step S20 provided in this embodiment of the invention includes:

[0038] Acquire the amplification band data of the starting strain at each site, and compare the amplification band data of each tested strain with the amplification band data of the starting strain site by site;

[0039] For each site, when the tested strain has an amplification band at the site but the originating strain does not, the mutation state of the site is marked as a new band.

[0040] When the test strain does not have an amplification band at the site but the starting strain exists, the mutation state at the site is marked as a deletion band;

[0041] When the amplification band status of the tested strain and the starting strain at the same site is consistent, the mutation status of the site is marked as no mutation has occurred;

[0042] The mutation status of all sites was arranged in order of site to construct the mutagenic differential fingerprint vector for each tested strain.

[0043] In step S10, amplification band data of the starting strain at all molecular marker sites were obtained. The starting strain was the original strain preserved before space-induced mutagenesis, and its amplification band data was obtained synchronously with that of the test strains in step S10. The amplification band data of each test strain were compared site-by-site with the amplification band data of the starting strain. For each site, the amplification band data of both strains were in a binary state: 1 indicates the presence of an amplification band at that site, and 0 indicates the absence of an amplification band. The comparison logic determined the mutation state marker based on the different combinations of these states.

[0044] When the amplification band data of the tested strain at the site is 1 while that of the originating strain is 0, it indicates that the tested strain produced a new amplification band at that site. This mutation state is marked as a new band to distinguish it from other deletion types. A new band corresponds to a genome insertion or repetitive sequence amplification event induced by space-induced mutagenesis. When the tested strain's amplification band at the site is 0 while that of the originating strain is 1, it indicates that the tested strain lost the original amplification band at that site. This mutation state is marked as a deletion band. A deletion band corresponds to a DNA fragment deletion or primer binding site mutation induced by mutagenesis. When the amplification band data of the tested strain and the originating strain are the same at the site, the mutation state is marked as no mutation occurred.

[0045] The mutation states of all 128 sites are arranged sequentially to construct the mutagenic differential fingerprint vector of the tested strain. Each bit of the mutagenic differential fingerprint vector records the mutation state of that site relative to the originating strain. New bands are represented by the number 1, missing bands by the number -1, and no mutation by the number 0. It is important to note that the genotype data used in the subsequent association analysis in step S40 is the original amplified band data of each tested strain at each site obtained in step S10, encoded as either 0 or 1. These two sets of data are used independently in subsequent analyses. For example, the mutagenic differential fingerprint vector of a certain tested strain has 128 sites, of which 5 sites (including sites 3 and 7) are 1, 11 sites (including sites 5 and 11) are -1, and the remaining 112 sites are 0.

[0046] The following technical effects were achieved through this step:

[0047] By constructing a mutagenic difference fingerprint vector through site-by-site difference comparison with the starting strain, the molecular variation of each tested strain relative to the starting strain is intuitively quantified into fingerprint sequences of three states: newly added bands, missing bands, and no variation. This provides molecular-level input features that directly reflect mutagenic differences for subsequent association analysis.

[0048] S30: Molecular marker detection is performed on multiple single colonies of the same test strain, the consistency index of the detection results at each site is calculated, and the genotype data of the test strain at the corresponding site is corrected according to the consistency index.

[0049] During subculture of mutant strains induced by space mutagenesis, the incomplete stability of somatic cell variation can lead to inconsistent amplification banding at certain loci in different single colonies of the same tested strain. This inconsistency introduces random errors into the genotype data. By performing molecular marker detection on multiple single colonies of the same tested strain and calculating the consistency index of each locus, unreliable loci with consistency below the threshold can be excluded from subsequent association analysis and breeding value scoring, ensuring the stability and reproducibility of the genotype data used.

[0050] Step S30 provided in this embodiment of the invention includes:

[0051] At least three single colonies were randomly selected from the same test strain, and molecular marker detection was performed on each single colony to obtain amplification band data of each single colony at each point.

[0052] For each site, the number of times the detection results are consistent in each single colony is counted, and the number of consistent results is divided by the total number of single colonies to obtain the site consistency index.

[0053] Genotype data of sites with a consistency index below a preset threshold are marked as pending confirmation and will not be included in subsequent association analysis and breeding value scoring calculations.

[0054] Genotype data of loci with a consistency index higher than or equal to a preset threshold are retained and used in subsequent association analysis and breeding value scoring calculations.

[0055] The specific implementation method is as follows:

[0056] From the original mycelial cultures of each tested strain, mycelial tips were randomly picked using an inoculation needle and transferred to new PDA plates. These plates were then incubated at 25°C until independent colonies formed. At least three single colonies were randomly selected from each tested strain. For example, three single colonies were collected. Genomic DNA was extracted from each single colony using the same CTAB method as in step S10. PCR amplification and electrophoresis were performed using the same ISSR primer set, and the amplified band data for each single colony at all 128 molecular marker sites were recorded.

[0057] For each site, the number of consistent results in each single colony is counted. Specifically, if 3 out of 3 single colonies of the tested strain show amplification bands (i.e., all single colonies show amplification bands) or 3 single colonies do not show amplification bands (i.e., all single colonies show no amplification bands), the consistency count is 3, and the consistency index = 3 ÷ 3 = 1.0. If 2 single colonies show amplification bands and 1 single colony shows no amplification band, the consistency count is 2, meaning the results with amplification bands are in the majority, and the consistency index = 2 ÷ 3 ≈ 0.67. If 2 single colonies show no amplification bands and 1 single colony shows amplification band, the consistency count is 2, meaning the results without amplification bands are in the majority, and the consistency index = 2 ÷ 3 ≈ 0.67. If the number of single colonies increases to an even number of 4 or more, it is possible that the results with and without amplification bands will each account for half, making it impossible to form a majority consensus. When a tie occurs, a representative genotype that constitutes the majority cannot be formed. The genotype data of this locus is not reproducible and is therefore excluded. The consistency index of this locus is set to 0, and it is marked as pending confirmation. Its genotype data will not be included in subsequent association analysis and breeding value scoring calculations.

[0058] The pre-set threshold for the consistency index is determined based on the number of single colonies and the acceptable reliability requirements of genotype data. For example, when the number of single colonies is 3, the threshold is 0.67, corresponding to more than two out of the three single colonies showing consistent results, indicating the existence of a majority representative genotype. This threshold is based on the majority consensus principle. When the number of single colonies takes other values, the pre-set threshold for the consistency index can be obtained by dividing the minimum number of consensuses required for majority consensus by the total number of single colonies. For example, the minimum number of consensuses required for majority consensus is equal to the total number of single colonies divided by 2, rounded down, and then added by 1.

[0059] Genotype data at loci with a consistency index below the threshold are marked as pending confirmation. This indicates that the genotype detection results of the tested strain at this locus show low consistency among individual colonies, and the genotype data has uncertainty. The genotype data for these loci is set to 0 by default, and these loci are not included in subsequent association analysis and breeding value scoring. Genotype data at loci with a consistency index at or above the threshold are retained and included in subsequent association analysis.

[0060] For example, after molecular marker detection at 128 loci for three single colonies of a certain tested strain, the homology index of 112 loci was 1.0, the homology index of 10 loci was 0.67, and the homology index of 6 loci was below 0.67. The 6 loci with a homology index below 0.67 were marked as pending confirmation, and the corrected genotype data of the remaining 122 loci were retained for subsequent analysis.

[0061] The following technical effects were achieved through this step:

[0062] By performing molecular marker detection on multiple single colonies of the same test strain and calculating the consistency index of each site, sites with consistency below the threshold were marked as pending confirmation and excluded from subsequent analysis. This eliminated the problem of unreliable genotype data caused by the instability of somatic cell variation after space mutagenesis, and improved the credibility of the genotype data on which subsequent association analysis and breeding value scoring depend.

[0063] S40: Using the corrected genotype data of each tested strain at each locus as the independent variable and the agronomic trait data as the dependent variable, perform locus-by-locus association analysis to screen out the core molecular marker combinations associated with the target agronomic traits and determine the contribution weight of each core marker locus.

[0064] High yield in *Dictyophora indicum* is a complex trait, with multiple traits such as mycelial growth rate, biological efficiency, fruiting body size, and stress resistance contributing to high yield to varying degrees. Furthermore, the phenotypic values ​​of the same trait fluctuate across different cultivation batches. Directly performing association analysis with equal weights on each trait fails to distinguish the differential contributions of different traits to the high-yield target, and traits with poor stability may introduce environmental noise. Therefore, it is necessary to first determine the weighting coefficients based on the contribution of each trait to the high-yield target, then adjust the weights based on the stability of the traits across different batches. The weighted and adjusted agronomical trait data should then be used as the dependent variable in the association analysis, ensuring that the selected core marker loci more accurately reflect their actual contribution to the overall high-yield target.

[0065] Step S40 provided in this embodiment of the invention includes:

[0066] S41: Weighting and stability correction of agronomic trait data, including:

[0067] Agronomic trait data for each tested strain were obtained. The agronomic trait data included measured values ​​for multiple trait types such as mycelial growth rate, biological efficiency, fruiting body size, and stress resistance. The fruiting body size included stipe length, stipe diameter, veil length, and veil thickness.

[0068] Based on the degree of contribution of each trait type to the high-yield breeding target, the weight coefficient of each trait type is determined, and the weight coefficient is positively correlated with the degree of contribution of the trait type to the high-yield target.

[0069] The measured values ​​of each tested strain in each trait type are multiplied by the weighting coefficient of the corresponding trait type to obtain the weighted agronomic trait data.

[0070] Obtain the measured values ​​of agronomic traits of the same test strain in at least three cultivation batches, and calculate the coefficient of variation of each trait type in different batches;

[0071] The stability coefficient of each trait type is determined based on the coefficient of variation, and the stability coefficient is negatively correlated with the coefficient of variation;

[0072] The weighting coefficients of each trait type are corrected based on the stability coefficient. The corrected weighting coefficients are equal to the original weighting coefficients multiplied by the stability coefficient. The corrected weighted agronomic trait data are used as the dependent variable for the site-by-site association analysis.

[0073] The specific implementation method is as follows:

[0074] Obtain the agronomic trait data of each tested strain obtained in step S10. Optionally, the agronomic trait data includes four trait types: mycelial growth rate, biological efficiency, fruiting body size, and stress resistance. Fruiting body size includes four indicators: stipe length, stipe diameter, veil length, and veil thickness. High-yield breeding targets require rapid mycelial growth to shorten the cultivation cycle, high biological efficiency to improve raw material conversion rate, moderate fruiting body size to meet commercial specifications, and strong stress resistance to reduce cultivation risks. Weighting coefficients are determined based on the contribution of each trait type to the high-yield breeding target. The weighting coefficients are dynamically determined by the measured agronomic trait data of the current batch of tested strains according to the proportion of variation contribution. First, the measured values ​​of the four traits were normalized for each strain to eliminate the influence of different unit dimensions for mycelial growth rate, biological efficiency, fruiting body size, and stress resistance. Then, the variation contribution of each trait's measured value among all tested strains was calculated. The variation contribution was measured as the sum of squared deviations of the normalized measured value of each strain from the population mean of that trait. The proportion of each trait's variation contribution to the total variation contribution of the four traits was used as the weighting coefficient for that trait. The larger the proportion, the higher the distinguishability of that trait among the tested strains, the greater its discriminative effect on distinguishing and selecting high-yielding strains, and the higher its corresponding weight. For example, in the current batch, the inter-plant variation contributions of the four traits—mycelial growth rate, biological efficiency, fruiting body size, and stress resistance—are 2.5, 3.0, 3.0, and 1.5, respectively, with a total variation contribution of 10. The normalized proportions, i.e., the weighting coefficients, are 0.25, 0.30, 0.30, and 0.15 for mycelial growth rate, biological efficiency, fruiting body size, and stress resistance, respectively, and the sum of the weighting coefficients is 1. Among them, the weight of 0.30 for fruiting body size is evenly distributed among the four sub-indicators: stipe length, stipe diameter, veil length, and veil thickness, each with a weight of 0.075.

[0075] The measured values ​​of each tested strain for each trait type are multiplied by the corresponding weighting coefficient to obtain the weighted agronomic trait data. For example, a tested strain has a mycelial growth rate of 1.5 mm / d, a biological efficiency of 12.5%, a stipe length of 128 mm, a stipe diameter of 18.5 mm, a veil length of 85 mm, a veil thickness of 2.8 mm, and a stress resistance of 85%. The weighted agronomic trait data = 1.5 × 0.25 + 12.5 × 0.30 + 128 × 0.075 + 18.5 × 0.075 + 85 × 0.075 + 2.8 × 0.075 + 85 × 0.15 = 0.375 + 3.750 + 9.600 + 1.388 + 6.375 + 0.210 + 12.750 = 34.448.

[0076] Obtain the measured values ​​of the above four traits for the same tested strain in three independent cultivation batches. Specifically, calculate the coefficient of variation for each trait type in different batches. The coefficient of variation is equal to the standard deviation of the measured values ​​of that trait in the three batches divided by the arithmetic mean. The standard deviation is calculated based on the sample standard deviation, i.e., the square root of the sum of the squares of the differences between the measured values ​​of each batch and the mean, divided by the batch number minus 1. For example, the mycelial growth rates of a certain tested strain in three batches were 1.5, 1.3, and 1.7 mm / d, with a mean of 1.5 mm / d and a standard deviation of 0.2 mm / d. The coefficient of variation is approximately 0.2 ÷ 1.5 ≈ 0.133. The biological efficiency in the three batches was 12.5%, 11.8%, and 13.2%, with a mean of 12.5% ​​and a standard deviation of 0.7%. The coefficient of variation is approximately 0.7 ÷ 12.5 = 0.056. The stipe lengths of the three batches were 128, 122, and 134 mm, with a mean of 128 mm, a standard deviation of 6 mm, and a coefficient of variation of 6 ÷ 128 ≈ 0.047. The stipe diameters of the three batches were 18.5, 17.8, and 19.2 mm, with a mean of 18.5 mm, a standard deviation of 0.7 mm, and a coefficient of variation of 0.7 ÷ 18.5 ≈ 0.038. The veil lengths of the three batches were 85, 80, and 90 mm, with a mean of 85 mm, a standard deviation of 5 mm, and a coefficient of variation of 5 ÷ 85 ≈ 0.059. The veil thicknesses of the three batches were 2.8, 2.5, and 3.1 mm, with a mean of 2.8 mm, a standard deviation of 0.3 mm, and a coefficient of variation of 0.3 ÷ 2.8 ≈ 0.107. The stress resistance of the three batches was 85%, 82%, and 88%, with a mean of 85%, a standard deviation of 3%, and a coefficient of variation of 3 ÷ 85 ≈ 0.035.

[0077] The stability coefficients for each trait type are determined based on the coefficient of variation. The stability coefficient = 1 - coefficient of variation; a stability coefficient is set to 0 when the coefficient of variation is greater than or equal to 1. The stability coefficients for mycelial growth rate, biological efficiency, stipe length, stipe diameter, veil length, stipe diameter, and stress resistance are as follows: 1 - 0.133 = 0.867; 1 - 0.056 = 0.944; 1 - 0.047 = 0.953; 1 - 0.038 = 0.962; 1 - 0.059 = 0.941; 1 - 0.107 = 0.893; and 1 - 0.035 = 0.965.

[0078] The weighting coefficients of each trait type were adjusted based on the stability coefficient. The adjusted weighting coefficients were equal to the original weighting coefficients multiplied by the stability coefficient. The adjusted weighting coefficients for mycelial growth rate were: 0.25 × 0.867 ≈ 0.217; biological efficiency: 0.30 × 0.944 ≈ 0.283; stipe length: 0.075 × 0.953 ≈ 0.071; stipe diameter: 0.075 × 0.962 ≈ 0.072; veil length: 0.075 × 0.941 ≈ 0.071; veil thickness: 0.075 × 0.893 ≈ 0.067; and stress resistance: 0.15 × 0.965 ≈ 0.145. The sum of the corrected weight coefficients = 0.217 + 0.283 + 0.071 + 0.072 + 0.071 + 0.067 + 0.145 = 0.926, which is no longer 1, so it is normalized. The normalized weighting coefficients for mycelial growth rate are 0.217 ÷ 0.926 ≈ 0.234, biological efficiency is 0.283 ÷ 0.926 ≈ 0.306, stipe length is 0.071 ÷ 0.926 ≈ 0.077, stipe diameter is 0.072 ÷ 0.926 ≈ 0.078, veil length is 0.071 ÷ 0.926 ≈ 0.077, veil thickness is 0.067 ÷ 0.926 ≈ 0.072, and stress resistance is 0.145 ÷ 0.926 ≈ 0.157. The measured agronomic traits of each tested strain were reweighted using these corrected and normalized weighting coefficients to obtain the corrected weighted agronomic trait data. For example, the corrected weighted agronomic trait data of the tested strain = 1.5×0.234+12.5×0.306+128×0.077+18.5×0.078+85×0.077+2.8×0.072+85×0.157=0.351+3.825+9.856+1.443+6.545+0.202+13.345=35.567.

[0079] By introducing multi-trait weighting and stability correction, the differentiated contributions of different traits to the high-yield target were quantified. The stability differences of traits in cultivation batches were included in the weight correction, making the dependent variable of the association analysis more accurately reflect the comprehensive agronomic trait level of each tested strain and improving the accuracy of the association analysis.

[0080] S42: Site-by-site association analysis and core marker screening, including:

[0081] For each molecular marker locus, the genotype data of the locus after correction on each tested strain were used as independent variables, and the agronomic trait data of the corresponding tested strain were used as dependent variables. Single-marker regression analysis was performed to obtain the regression coefficient and significance probability value of the locus.

[0082] Loci with a significance probability value less than a preset significance level are retained as candidate marker loci;

[0083] Based on the absolute value of the regression coefficients of each candidate marker site, they are sorted in descending order, and the top N sites are selected as the core molecular marker combination.

[0084] The regression coefficients of each core marker site are used as the contribution weights of the site.

[0085] The specific implementation method is as follows:

[0086] Using the genotype data of each tested strain at each locus after correction in step S30 as independent variables and the corrected weighted agronomic trait data as dependent variables, single-label regression analysis was performed at each locus. That is, for each locus, a univariate linear regression was performed with the genotype data of that locus as the unique independent variable and the corrected weighted agronomic trait data as the dependent variable.

[0087] For each molecular marker site, a regression equation Y is established. j =β 0j +β 1j ×X j +ε j , where Y j For the corrected weighted agronomic traits of the tested strains, X j For the genotypic data of the tested strain at this locus, β 0j For the regression intercept, β 1j ε is the regression coefficient. j This represents random error. The regression coefficient β 1j This reflects the weighted change in agronomic traits corresponding to each unit change in the genotype at that locus. The significance of the regression is determined by the F-test, with the significance probability being the overall significance p-value of the regression equation. The smaller the p-value, the more significant the association between that locus and the agronomic trait.

[0088] Preferably, the preset significance level is determined based on commonly used thresholds in molecular marker association analysis, and is set to 0.05. Loci with p-values ​​less than 0.05 are retained as candidate marker loci, as these loci have a statistically significant association with the adjusted weighted agronomic traits. Loci with p-values ​​greater than or equal to 0.05 are excluded because they have no significant association with the agronomic traits. For example, after single-marker regression analysis of 128 loci, 48 loci had p-values ​​less than 0.05 and were retained as candidate marker loci.

[0089] Forty-eight candidate marker loci were sorted from largest to smallest based on the absolute value of their regression coefficients. A larger absolute value indicates a greater impact of genotypic changes on agronomic traits and a more significant contribution to high-yield breeding. The top N loci were selected as the core molecular marker combination. The value of N was determined based on the cumulative contribution of the absolute values ​​of the regression coefficients of the candidate marker loci: the absolute values ​​of the regression coefficients were accumulated sequentially from largest to smallest. When the accumulated value reached a preset proportion of the total absolute values ​​of the regression coefficients of the candidate marker loci, the number of loci reaching this proportion was taken as N. The preset proportion was 80%, meaning that loci contributing the vast majority of the overall agronomic traits were retained. The lower limit of N was set to 5 to ensure that the core combination retained a sufficient number of multi-gene effect loci, and the upper limit was set to 20 to avoid introducing noise loci with excessively small absolute values ​​of regression coefficients and negligible marginal contributions, while also controlling the number of loci for subsequent cultivation validation and weight adjustment. For example, after the 48 candidate marker loci are sorted in descending order of the absolute value of their regression coefficients, the sum of the absolute values ​​of the regression coefficients of the top 10 loci accounts for 83% of the sum of all 48 loci, reaching the preset cumulative contribution ratio of 80%. Therefore, N is set to 10, and the top 10 loci are selected as the core molecular marker combination. The regression coefficient of each core marker locus is the contribution weight of that locus. A positive contribution weight indicates that the presence of an amplification band at that locus is conducive to high yield, while a negative contribution weight indicates that the presence of an amplification band at that locus is not conducive to high yield.

[0090] For example, among the top 10 core marker loci, the regression coefficient of locus ISSR-047 is +1.85, indicating that when an amplification band is present at this locus, the corrected weighted agronomic trait data improves by an average of 1.85 units, making a positive contribution to high yield. The regression coefficient of locus ISSR-086 is -1.62, indicating that when an amplification band is present at this locus, the corrected weighted agronomic trait data decreases by an average of 1.62 units, making a negative contribution to high yield. The contribution weight of each core marker locus is determined by the positive or negative value of the regression coefficient, and is used for subsequent breeding value scoring calculations.

[0091] The following technical effects were achieved through this step:

[0092] Weighting coefficients were determined based on the contribution of each agronomic trait to the high-yield breeding objective, and stability corrections were made using the coefficient of variation among cultivation batches to ensure that the dependent variable in the association analysis more accurately reflects the comprehensive agronomic trait level of each tested strain. Single-marker regression analysis was used to screen for core marker loci combinations significantly associated with weighted agronomic traits, and the regression coefficients of each core marker locus were used as contribution weights to establish a quantitative association between molecular marker genotypic differences and the comprehensive high-yield objective. This provided marker loci and their quantitative weights that matched the breeding objective for subsequent breeding value scoring.

[0093] S50: Calculate the breeding value score of each tested strain based on its genotype and corresponding contribution weight at the core marker site, and screen candidate strains based on the breeding value score.

[0094] Step S40 has identified core marker loci and determined their contribution weights. By converting the genotypes of each tested strain at each core marker locus into weighted scores and summing them, a unified breeding value score is quantified. Candidate strains are then selected for cultivation validation based on their scores. After cultivation validation of the candidate strains, the contribution weights are adjusted based on the deviation between the measured agronomic traits and the predicted scores, as well as the stability of the regression coefficients at each core marker locus. An incremental learning mechanism is used to continuously optimize the contribution weights as the breeding process progresses.

[0095] Step S50 provided in this embodiment of the invention includes:

[0096] S51: Calculate breeding value scores and screen candidate strains, including:

[0097] For each tested strain, the genotype data of the tested strain at each core marker site is extracted. The genotype data is 0 or 1, where 1 indicates the presence of an amplification band and 0 indicates the absence of an amplification band.

[0098] The genotype data of each core marker locus is multiplied by the contribution weight of that locus to obtain the weighted score of each core marker locus.

[0099] The breeding value score of the tested strain is obtained by summing the weighted scores of all core marker sites.

[0100] The specific implementation method is as follows:

[0101] Genotype data were extracted from the 10 core marker loci obtained in step S40 for each tested strain. For loci marked as pending confirmation in step S30, if the locus belongs to a combination of core marker loci, its genotype data was treated as 0 in the breeding value score calculation, and the weighted score of the locus was also counted as 0 in the sum. This method ensures that the total number of loci involved in the breeding value score calculation is the same for each tested strain, and the scoring results are comparable among different tested strains.

[0102] The genotype data for each core marker locus is multiplied by its contribution weight to obtain a weighted score. A positive contribution weight indicates that the presence of an amplification band at that locus is conducive to high yield; a genotype of 1 results in a positive weighted score, and a genotype of 0 results in a zero weighted score. Conversely, a negative contribution weight indicates that the presence of an amplification band at that locus is detrimental to high yield; a genotype of 1 results in a negative weighted score, and a genotype of 0 results in a zero weighted score. The weighted scores of all 10 core marker loci are summed to obtain the breeding value score of the tested strain. A higher breeding value score indicates a higher degree of aggregation of beneficial alleles at the core marker loci, and a greater expected high-yield potential.

[0103] For example, a certain tested strain has genotype 1 at 7 out of 10 core marker loci and genotype 0 at 3 loci. Among the 7 loci with genotype 1, 4 loci have positive contribution weights, with a total weighted score of +5.82; 3 loci have negative contribution weights, with a total weighted score of -2.75. The breeding value score of this tested strain = 5.82 + (-2.75) = 3.07.

[0104] All tested strains were ranked from highest to lowest according to their breeding value score, and the top few strains with the highest scores were selected as candidate strains. For example, the breeding value scores of the 36 tested strains ranged from -1.68 to 5.23, and the top 5 strains with the highest scores were selected as candidate strains for subsequent cultivation verification.

[0105] S52: Feedback correction and incremental learning of contribution weights, including:

[0106] The selected candidate strains were cultivated for verification, and the measured agronomic traits of the candidate strains in the cultivation verification were obtained.

[0107] Calculate the measured breeding value score for each candidate strain, compare the measured breeding value score with the predicted breeding value score, and calculate the prediction deviation.

[0108] Obtain the regression coefficients of each core marker locus and the target agronomic trait in at least three cultivation batches, calculate the coefficient of variation of the regression coefficients of each core marker locus in different batches, and determine the association stability coefficient of each core marker locus based on the coefficient of variation, wherein the association stability coefficient is negatively correlated with the coefficient of variation.

[0109] The contribution weights of each core marker site are adjusted based on the prediction bias and the association stability coefficient.

[0110] Obtain the breeding value score sequence of candidate strains in multiple consecutive cultivation batches, and calculate the moving average of the breeding value score sequence;

[0111] When the moving average is greater than the preset termination threshold, and the change in the moving average in two or more consecutive batches is less than the preset fluctuation threshold, the correction of the contribution weight of each core marker site is stopped, and the current contribution weight is used as the final weight for the calculation of the breeding value score of subsequent tested strains.

[0112] The molecular marker loci genotype data and agronomic trait data of each tested strain were obtained as historical training data.

[0113] The measured agronomic traits and corresponding molecular marker locus genotype data of the candidate strains in cultivation verification were obtained as new batches of verification data.

[0114] The degree of difference in genotype frequency distribution between each core marker locus in the new batch of validation data and in the historical training data is calculated as the distribution offset of the locus.

[0115] The fusion weight of a site is determined based on the distribution offset of each core marker site, and the fusion weight is negatively correlated with the distribution offset.

[0116] Based on the fusion weights, historical training data and new batch of validation data are weighted and sampled to construct an incremental training set;

[0117] The contribution weights of each core marker site are updated using the incremental training set.

[0118] The specific implementation method is as follows:

[0119] Specifically, the candidate strains are cultivated and verified in at least three batches. The measured values ​​of each candidate strain in each batch are obtained in terms of four traits: mycelial growth rate, biological efficiency, fruiting body size and stress resistance. The measured breeding value score of each candidate strain is calculated using the same weighted and stability correction method as in step S40.

[0120] The measured breeding value score is compared with the predicted breeding value score, and the prediction bias is calculated. The prediction bias equals the absolute value of the measured value minus the predicted value, divided by the predicted value. For example, if the predicted breeding value score of a candidate strain is 4.52, and the measured breeding value scores of three batches of cultivation verification are 4.12, 4.35, and 4.28, with a mean of 4.25, then the prediction bias = |4.25 - 4.52| ÷ 4.52 ≈ 0.060, or 6.0%.

[0121] Obtain the regression coefficients of each of the 10 core marker loci and the weighted agronomic traits in three cultivation batches. Specifically, calculate the coefficient of variation (COP) of the regression coefficient for each core marker locus across the three batches. The COP is equal to the standard deviation of the regression coefficient divided by its absolute mean. The standard deviation is also calculated using the sample standard deviation, consistent with the calculation method for the coefficient of variation of agronomic traits in step S40. For example, the regression coefficients for locus ISSR-047 in the three batches are 1.85, 1.72, and 1.78, with an absolute mean of 1.783 and a sample standard deviation of 0.065. The COP = 0.065 ÷ 1.783 ≈ 0.036. The regression coefficients for locus ISSR-086 in the three batches are -1.62, -1.15, and -2.08, with an absolute mean of 1.617 and a sample standard deviation of 0.465. The COP = 0.465 ÷ 1.617 ≈ 0.288.

[0122] The association stability coefficient is determined based on the coefficient of variation. The association stability coefficient = 1 - coefficient of variation, with a value of 0 when the coefficient of variation is greater than or equal to 1. The association stability coefficient for locus ISSR-047 is approximately 0.964 (1 - 0.036), and for locus ISSR-086 it is approximately 0.712 (1 - 0.288). A higher association stability coefficient (closer to 1) indicates greater stability of the genetic effect of that locus across different batches.

[0123] The contribution weights of each core marker locus are adjusted based on the prediction bias and the association stability coefficient of each core marker locus. The adjusted contribution weight = original contribution weight × (1 - prediction bias) × association stability coefficient. For example, the original contribution weight of locus ISSR-047 is +1.85, the prediction bias is 0.060, and the association stability coefficient is 0.964. The adjusted contribution weight = 1.85 × (1 - 0.060) × 0.964 = 1.85 × 0.940 × 0.964 ≈ 1.677. The original contribution weight of locus ISSR-086 is -1.62, and the adjusted contribution weight = -1.62 × 0.940 × 0.712 ≈ -1.084. The adjusted contribution weights shrink towards zero. Locuses with more stable locus effects and smaller prediction biases experience smaller shrinkage, retaining more of the original information. Locuses with large fluctuations in locus effects and large prediction biases experience larger shrinkage, significantly weakening their weights.

[0124] Obtain the breeding value score sequence of candidate strains from multiple consecutive cultivation batches. After each cultivation batch, recalculate the breeding value score of each candidate strain using the corrected contribution weight of that batch, and take the arithmetic mean of the breeding value scores of the most recent three batches as the moving average. When the moving average is greater than a preset termination threshold and the change between two consecutive batches is less than a preset fluctuation threshold, it indicates that the breeding value score has stabilized at the breeding target level, and the correction of the contribution weight is stopped, with the current contribution weight being used as the final weight.

[0125] The preset termination threshold is the maximum value of the measured breeding value score of each candidate strain in the previous round of cultivation verification. This value represents the highest score level actually achieved in the current breeding stage. When the moving average exceeds this threshold, it indicates that the breeding value score has stabilized above the current optimal level, and the marginal benefit of continuing iteration decreases. The preset fluctuation threshold is 5%, which is based on the commonly used tolerance for the stability of genetic effect estimation in breeding practice: when the score change between adjacent batches is less than 5%, the fluctuation of the contribution weight of the core marker site has no substantial impact on the ranking of the breeding value scores of each tested strain, and the score can be determined to have converged. For example, the highest measured breeding value score in the previous round of cultivation verification was 4.35, and the preset termination threshold was 4.35; the moving average was 4.38, which is greater than 4.35; the changes between two consecutive batches were 2.1% and 1.8%, respectively, both less than 5%, which met the iteration termination condition, and the correction of the contribution weight was stopped.

[0126] The molecular marker locus genotype data and agronomic trait data of each tested strain were used as historical training data, while the measured agronomic trait data and corresponding molecular marker locus genotype data of the candidate strains in cultivation validation were used as the new batch of validation data. The degree of difference in genotype frequency distribution of each core marker locus between the new batch of validation data and the genotype frequency distribution in the historical training data was calculated as the distribution offset of the locus.

[0127] Preferably, the distribution shift is calculated using KL divergence. The frequencies of genotype 0 and 1 at the locus in the two data sets are used as discrete distribution probabilities to calculate the information divergence value of the new batch distribution relative to the historical distribution. Specifically, the formula for calculating KL divergence is: KL(P new ||P hist )=P new(0) ×ln(P new(0) ÷P hist(0) )+P new(1) ×ln(P new(1) ÷P hist(1) ), where P hist(0) With P hist(1)P represents the frequency of genotype 0 and genotype 1 for this core marker locus in historical training data, respectively. new(0) With P new(1) Let P and L represent the frequencies of genotype 0 and genotype 1 for this core marker locus in the new batch of validation data, respectively, where ln is the natural logarithm function. hist(0) or P hist(1) When P is zero, the frequency is replaced with a preset minimum probability value ε = 0.01 to avoid division by zero errors, while maintaining the sum of frequencies at 1. new(0) or P new(1) When the value is zero, the term is zero.

[0128] The fusion weight is determined based on the distribution offset of each core marker locus. Fusion weight = 1 - distribution offset; when the distribution offset is greater than or equal to 1, the fusion weight is 0. A smaller distribution offset indicates that the genotype distribution of the locus is closer in the old and new batches of data, the higher the reliability of the historical data, and the closer the fusion weight is to 1. A larger distribution offset indicates a significant change in the genotype distribution, a lower reliability of the historical data, and a smaller fusion weight.

[0129] Weighted sampling is performed on historical training data and the new batch of validation data based on fusion weights to construct an incremental training set. Preferably, the weighted sampling method is as follows: the total sampling size is set to M, and the value of M is determined comprehensively based on computational resources and model stability requirements; for example, it can be set to 80% of the historical training data sample size. The sampling size of the historical training data is M multiplied by the fusion weight of that locus, and the sampling size of the new batch of validation data is M multiplied by (1 minus the fusion weight of that locus). Sampling is performed with replacement, randomly drawing samples from both the historical training data and the new batch of validation data according to the corresponding sampling size, and merging the drawn samples to form the incremental training set for that locus. The closer the fusion weight is to 1, the higher the proportion of historical training data in the incremental training set, and the more the model relies on historically accumulated knowledge; the closer the fusion weight is to 0, the higher the proportion of the new batch of validation data, and the more the model focuses on recent validation results. Sampling of each core marker locus is performed independently, generating an incremental training set corresponding to each position, which is used for subsequent updates of the contribution weights of each locus.

[0130] For example, assuming a total sampling size M=100 and a fusion weight of 0.7 for a certain locus, 70 samples are randomly selected with replacement from the historical training data, and 30 samples are randomly selected with replacement from the new batch of validation data. These samples are then merged to form the incremental training set for that locus. The above process is repeated for each sampling point at each core marker locus. The regression analysis in step S40 is then re-executed using the incremental training set to update the contribution weights of each core marker locus, completing the incremental learning.

[0131] The following technical effects were achieved through this step:

[0132] By weighted summation of contribution weights, the genotypic differences of each tested strain at the core marker locus are quantified into a unified breeding value score, enabling multi-gene aggregation breeding objectives to be measured and ranked using a single numerical value. The contribution weights are corrected through cultivation validation feedback, and unstable locus effects are mitigated using prediction bias and association stability coefficients, making the corrected contribution weights more closely reflect actual cultivation performance. An incremental learning mechanism updates the contribution weights after each new batch of validation data accumulates, allowing the breeding value scoring model to continuously optimize as the breeding process progresses.

Claims

1. A molecular marker-assisted method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum*, characterized in that, include: Acquire amplification band data of each tested strain at multiple molecular marker sites, as well as agronomic trait data of each tested strain; The amplification band data of each tested strain were compared with the amplification band data of the starting strain at each site to construct the mutagenic differential fingerprint vector of each tested strain. Molecular marker detection was performed on multiple single colonies of the same test strain, and the consistency index of the detection results at each site was calculated. The genotype data of the test strain at the corresponding site was corrected based on the consistency index. Using the corrected genotype data of each tested strain at each locus as independent variables and the agronomic trait data as dependent variables, a locus-by-locus association analysis was performed to screen out the core molecular marker combinations associated with the target agronomic traits and determine the contribution weight of each core marker locus. Based on the genotype and corresponding contribution weight of each tested strain at the core marker site, the breeding value score of each tested strain is calculated, and candidate strains are screened based on the breeding value score.

2. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* var. *red* according to claim 1, characterized in that, The amplification band data of each tested strain were compared with the amplification band data of the starting strain at each site to construct the mutagenic differential fingerprint vector of each tested strain, including: Acquire the amplification band data of the starting strain at each site, and compare the amplification band data of each tested strain with the amplification band data of the starting strain site by site; For each site, when the tested strain has an amplified band at the site but the originating strain does not, the mutation state of the site is marked as a new band. When the test strain does not have an amplification band at the site but the starting strain exists, the mutation state at the site is marked as a deletion band; When the amplification band status of the tested strain and the starting strain at the same site is consistent, the mutation status of the site is marked as no mutation has occurred; The mutation status of all sites was arranged in order of site to construct the mutagenic differential fingerprint vector for each tested strain.

3. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* var. *red* according to claim 1, characterized in that, Molecular marker detection was performed on multiple single colonies of the same tested strain, and the consistency index of the detection results at each locus was calculated. The genotype data of the tested strain at the corresponding locus was then corrected based on the consistency index, including: At least three single colonies were randomly selected from the same test strain, and molecular marker detection was performed on each single colony to obtain amplification band data at each point of each single colony; For each site, the number of times the detection results are consistent in each single colony is counted, and the number of consistent results is divided by the total number of single colonies to obtain the site consistency index. Genotype data of sites with a consistency index below a preset threshold are marked as pending confirmation and will not be included in subsequent association analysis and breeding value scoring calculations. Genotype data of loci with a consistency index higher than or equal to a preset threshold are retained and used in subsequent association analysis and breeding value scoring calculations.

4. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* with molecular marker assistance according to claim 1, characterized in that, Core molecular marker combinations associated with target agronomic traits were screened, and the contribution weight of each core marker site was determined, including: For each molecular marker locus, the genotype data of the locus after correction on each tested strain were used as independent variables, and the agronomic trait data of the corresponding tested strain were used as dependent variables. Single-marker regression analysis was performed to obtain the regression coefficient and significance probability value of the locus. Loci with a significance probability value less than a preset significance level are retained as candidate marker loci; Based on the absolute value of the regression coefficients of each candidate marker site, they are sorted in descending order, and the top N sites are selected as the core molecular marker combination. The regression coefficients of each core marker site are used as the contribution weights of the site.

5. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* with molecular marker assistance according to claim 1, characterized in that, Based on the genotype and corresponding contribution weight of each tested strain at the core marker site, the breeding value score of each tested strain was calculated, including: For each tested strain, the genotype data of the tested strain at each core marker site is extracted. The genotype data is 0 or 1, where 1 indicates the presence of an amplification band and 0 indicates the absence of an amplification band. The genotype data of each core marker locus is multiplied by the contribution weight of that locus to obtain the weighted score of each core marker locus. The breeding value score of the tested strain is obtained by summing the weighted scores of all core marker sites.

6. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* with molecular marker assistance according to claim 1, characterized in that, Also includes: Agronomic trait data for each tested strain were obtained. The agronomic trait data included measured values ​​for multiple trait types such as mycelial growth rate, biological efficiency, fruiting body size, and stress resistance. The fruiting body size included stipe length, stipe diameter, veil length, and veil thickness. Based on the degree of contribution of each trait type to the high-yield breeding target, the weight coefficient of each trait type is determined, and the weight coefficient is positively correlated with the degree of contribution of the trait type to the high-yield target. The measured values ​​of each tested strain in each trait type are multiplied by the weighting coefficient of the corresponding trait type to obtain weighted agronomic trait data. The weighted agronomic trait data are then used as the dependent variable for the site-by-site association analysis.

7. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* var. *red* according to claim 6, characterized in that, Also includes: Obtain the measured values ​​of agronomic traits of the same test strain in at least three cultivation batches, and calculate the coefficient of variation of each trait type in different batches; The stability coefficient of each trait type is determined based on the coefficient of variation, and the stability coefficient is negatively correlated with the coefficient of variation; The weighting coefficients of each trait type are adjusted according to the stability coefficient, and the adjusted weighting coefficients are equal to the original weighting coefficients multiplied by the stability coefficient.

8. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* var. *red* according to claim 7, characterized in that, Also includes: Obtain the breeding value score sequence of candidate strains in multiple consecutive cultivation batches, and calculate the moving average of the breeding value score sequence; When the moving average is greater than the preset termination threshold, and the change in the moving average in two or more consecutive batches is less than the preset fluctuation threshold, the correction of the contribution weight of each core marker site is stopped, and the current contribution weight is used as the final weight for the calculation of the breeding value score of subsequent tested strains.

9. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* with molecular marker assistance according to claim 1, characterized in that, Candidate strains were screened based on the breeding value score, followed by: The selected candidate strains were cultivated for verification, and the measured agronomic traits of the candidate strains in the cultivation verification were obtained. Calculate the measured breeding value score for each candidate strain, compare the measured breeding value score with the predicted breeding value score, and calculate the prediction deviation. Obtain the regression coefficients of each core marker locus and the target agronomic trait in at least three cultivation batches, calculate the coefficient of variation of the regression coefficients of each core marker locus in different batches, and determine the association stability coefficient of each core marker locus based on the coefficient of variation, wherein the association stability coefficient is negatively correlated with the coefficient of variation. The contribution weights of each core marker site are adjusted based on the prediction bias and the association stability coefficient.

10. The method for targeted breeding and selection of high-yielding strains of *Dictyophora indicum* var. *red* according to claim 9, characterized in that, Also includes: The molecular marker loci genotype data and agronomic trait data of each tested strain were obtained as historical training data. The measured agronomic traits and corresponding molecular marker locus genotype data of the candidate strains in cultivation verification were obtained as new batches of verification data. The degree of difference in genotype frequency distribution between each core marker locus in the new batch of validation data and in the historical training data is calculated as the distribution offset of the locus. The fusion weight of a site is determined based on the distribution offset of each core marker site, and the fusion weight is negatively correlated with the distribution offset. Based on the fusion weights, historical training data and new batch of validation data are weighted and sampled to construct an incremental training set; The contribution weights of each core marker site are updated using the incremental training set.

Images