Bark ratio prediction method

Abstract

To provide a method for predicting the bark thickness of a plant, that is, the bark ratio (or trunk ratio).SOLUTION: The present invention provides a method for predicting the bark ratio of a plant to be tested. The method includes: (A) obtaining genetic variation information of the plant to be tested that is a woody plant; and (B) obtaining a predicted value of the bark ratio of the plant to be tested based on an analytical model formula obtained by statistically analyzing the correlation between the genetic variation information of at least two closely related plants of the plant to be tested and the bark ratio of each closely related plant from the genetic variation information of the plant to be tested.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method for predicting bark ratio. [Background technology]

[0002] Afforestation projects take many years from planting to felling and harvesting, and it is important to efficiently select and plant trees with excellent growth potential and timber quality according to the intended use.

[0003] Wood is broadly divided into bark and xylem, and although the bark is sometimes used for building materials, the xylem is often the main component. For example, in papermaking applications, the xylem of wood is used, and the bark is removed by debarking prior to the pulp production process. This is because if bark is mixed into wood chips, problems such as a decrease in pulp yield and an increase in the amount of chemicals required will occur.

[0004] On the other hand, wood chips, a wood use, have recently been increasingly used as fuel for wood biomass resources. Wood chips are required to be produced at low cost compared to papermaking, so the bark removal process is generally not carried out. Bark contains ash and salts (Na, K, Cl) that have a negative effect on boilers, so it is considered preferable to have as little bark as possible mixed in.

[0005] Thus, wood with thin bark is required for papermaking and lumber applications, but evaluation of bark thickness is currently performed after the tree has fully grown, and a method for early selection of plantation trees with a low bark ratio is desired.

[0006] Non-Patent Document 1 describes that QTL analysis of eucalyptus revealed that quantitative trait loci (QTL) (RAPD markers) related to bark thickness are present on chromosome 5. Non-Patent Document 2 reports that QTL analysis was performed on poplar, and that 11 SNP markers were identified as QTL (RAPD markers) related to bark thickness. [Prior art documents] [Non-patent literature]

[0007] [Non-Patent Document 1] D. Grattapaglia et al. (1996) Genetics 144:1205-1214 https: / / pubmed.ncbi.nlm.nih.gov / 8913761 / [Non-patent document 2] R.Bdeir et al.BMC Plant Biology (2017)17:224 https: / / bmcplantbiol.biomedcentral.com / articles / 10.1186 / s12870-017-1166-4 Summary of the Invention [Problem to be solved by the invention]

[0008] Non-Patent Documents 1 and 2 were not studies aimed at prediction, and the number of QTLs related to bark thickness described in each is small. Furthermore, even if they were to be used as a prediction method, there would be problems such as differences in the correspondence with phenotypes depending on the habitat of the individual, and they cannot be said to be reliable prediction methods. The present invention aims to provide a method for predicting the bark thickness of a plant, i.e., the bark ratio (or trunk ratio). [Means for solving the problem]

[0009] The present invention provides the following: [1] (A): Obtaining genetic variation information for test plants that are woody plants; and (B): Obtain a predicted value for the bark ratio of the test plant based on an analytical model formula obtained by statistically analyzing the correlation between the genetic variation information of at least two closely related plants of the test plant and the bark ratio of each closely related plant from the genetic variation information of the test plant. Including, A method for predicting the bark percentage of a test plant. [2] The method according to [1], wherein the closely related plant is a plant having a diameter at breast height of 10 cm or more. [3] The method according to [1] or [2], wherein the statistical analysis is performed using a method selected from the group consisting of rrBLUP, GBLUP, GAUSS, Random Forest, RKHS, Ridge Regression, Lasso, Elastic Net, Bayesian Ridge Regression, Bayesian Lasso, BayesA, BayesB, and BayesC. [4] The method according to any one of [1] to [3], wherein the closely related plants include plants selected from plants of the same genus as the test plant but of different varieties or lineages. [5] The method according to claim 1 or 2, wherein the genetic variation information is information regarding a single nucleotide polymorphism. [6] The method according to [5], wherein the number of single nucleotide polymorphisms is 500 or more. [7] The method according to [5], wherein the number of single nucleotide polymorphisms is 1,000 to 47,000. [8] The method according to any one of [1] to [7], wherein the test plant is a Eucalyptus plant. [9] A method for selecting plants for wood chips, comprising obtaining predicted bark ratios for at least two test plants by the method described in any one of [1] to [7], and selecting test plants with low predicted bark ratios.

[10] The method according to [9], wherein the wood chips are for papermaking or fuel.

[11] A method for producing seedlings, comprising producing seedlings of plants selected by the selection method described in [9].

[12] A method for producing wood chips, comprising growing trees from seedlings produced by the production method described in

[11] , and processing the wood obtained by harvesting the trees into wood chips.

[13] The method according to

[12] , wherein the wood chips are for fuel.

[14] The method according to

[12] , further comprising peeling the wood or wood chips.

[15] The method according to

[14] , wherein the wood chips are for papermaking. [Effects of the Invention]

[0010] According to the present invention, the bark and stem ratios of plants can be determined and predicted in advance based on genomic information, allowing for the early selection of individuals with the appropriate bark and stem ratios for specific applications. For example, by selecting and breeding individuals predicted to have a low bark ratio, it is possible to harvest wood with a low bark ratio, which can be used as a raw material for wood chips for papermaking or fuel, thereby improving production efficiency. In other words, in papermaking applications, it is possible to increase chip productivity per unit area and reduce the cost of the bark removal process. In fuel applications, it is possible to reduce the amount of ash and salts (e.g., Na, K, Cl) contained in large amounts in bark, thereby reducing the adverse effects on boilers. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a diagram showing the flow of predicting the bark ratio (stem ratio) in the example. [Figure 2] FIG. 2 is an image diagram showing an example of correspondence information between samples and SNPs in the example. [Figure 3] Figure 3 shows the prediction accuracy of stem rate from prediction samples classified by DBH or randomly. [Figure 4] FIG. 4 is a diagram showing the prediction accuracy of test samples using prediction formulas created from prediction samples classified by DBH or randomly. DETAILED DESCRIPTION OF THE INVENTION

[0012] The method for predicting the bark ratio of the present invention includes the following (A) and (B): (A): Obtaining genetic variation information for woody plants; and (B): Obtain a predicted value of the bark ratio derived from the test plant based on an analytical model formula obtained by statistically analyzing the correlation between genetic variation information and bark ratio in at least two closely related plants of the test plant from the genetic variation information of the test plant.

[0013] <(A): Obtaining genetic variation information of test plants> [Test plants] The test plants are woody plants, including so-called forest tree species, for which predicted bark ratio (stem ratio) values ​​are to be obtained. Examples of woody plants include Eucalyptus plants (e.g., Eucalyptus uro-grandis, Eucalyptus urophylla, Eucalyptus pellita, Eucalyptus grandis, Eucalyptus brassiana), Corymbia plants, Cryptomeria plants (e.g., Cryptomeria japonica), Pinus plants, Prunus plants (e.g., Prunus spp., Prunus mume, Prunus tomentosa), Mangifera plants (e.g., Mangifera indica), Acacia plants, Myrica plants, Quercus plants (e.g., Quercus acutissima, etc.), Vitis plants, Malus plants, Rosa plants, Camellia plants, Jacaranda plants (e.g., Jacaranda mimosifolia, etc.), Persea plants (e.g., Persea americana, etc.), Chamaecyparis plants, Larix plants, Abies plants, Quercus plants, and Cunninghamia plants (e.g., Cunninghamia lanceolata, etc.). Among these, Eucalyptus plants, Corymbia plants, Cryptomeria plants, and Pinus plants are preferred, with Eucalyptus plants being more preferred.

[0014] [Genetic mutation information] Genetic variation information is information related to genetic variation. Genetic variation typically refers to DNA sequences that reflect individual differences among the reference plants constituting the reference plant group. Examples include single nucleotide polymorphisms (SNPs), restriction fragment length polymorphisms (RFLPs), repetitive sequences (satellite DNA, minisatellites, microsatellites), base deletions and additions, and amino acid substitutions, insertions, and deletions. Single nucleotide polymorphisms are preferred. If the genetic variation among the reference plants is known, it can be utilized; if it is unknown, it can be analyzed. Genetic variation can be obtained by genetic analysis of each reference plant. The genetic analysis method is not particularly limited, and examples include methods using DNA chips and next-generation sequencers. When the reference plants include Eucalyptus plants, SNP chips such as the Axiom® Eucalyptus Genotyping Array (developed by EMBRAPA, a Brazilian agricultural and livestock company) can be used. The Axiom Eucalyptus Genotyping Array was created based on SNPs selected from approximately 47 million SNPs obtained from a total of 240 individuals from 12 Eucalyptus species and one Corymbia species. The Axiom Eucalyptus Genotyping Array contains approximately 72,000 SNPs. It is an SNP chip that can be used with a wide range of Eucalyptus species. Individual differences in genetic variation do not need to be known to be associated with bark ratio or bark thickness.

[0015] The number of genetic variations is not particularly limited. When the genetic variations are single nucleotide polymorphisms, the number of single nucleotide polymorphisms is preferably 500 or more, more preferably 1,000 to 47,000, and even more preferably 5,000 to 15,000, from the viewpoint of improving prediction accuracy. It is preferable that the genetic variations used be those that have a high influence (marker effect) on the rooting rate per genetic variation obtained after statistical analysis. The influence of each genetic variation on the bark ratio can be calculated by statistically analyzing the SNP information of the test plant and the bark ratio actually measured for the test plant. Examples of statistical analysis methods will be described later.

[0016] Examples of genetic variation information for a test plant include the presence or absence of a specific genetic variation and the number of genetic variations. The genetic variation information is preferably quantified (scored).

[0017] <(B): Obtaining a predicted value for the bark ratio of the test plant based on an analytical model formula obtained by statistically analyzing the correlation between genetic variation information and bark ratio in at least two closely related plants of the test plant from the genetic variation information of the test plant> [Related plants] Related plants are used to clarify the correlation between genetic variation information and bark ratio. At least two related plants, preferably 10 or more, 15 or more, 20 or more, or 25 or more, are used to form a so-called training population. Related plants are usually selected from plants of the same or closely related species as the test plant, preferably plants of the same genus, and preferably two or more lineages or varieties of the same species. The number of reference plants is usually at least two, but a population as diverse as possible is preferable. Related plants are not particularly limited as long as they have a trunk and bark, but their diameter at breast height (DBH) is usually 6 cm or more, preferably 8 cm or more, more preferably 9 cm or more, 10 cm or more, or 11 cm or more. Having the DBH diameter within the above range can further improve prediction accuracy. It is presumed that the reason for this improved prediction accuracy is that individuals with growth suppressed by environmental influences can be excluded, allowing for a stronger capture of genetic influences. The upper limit is not particularly limited, and can be, for example, 30 cm or less or 25 cm or less. The related plant is preferably at least one year old, more preferably at least two years old, and even more preferably at least three years old. There is no particular upper limit, but for example, it may be 12 years old or younger, 10 years old or younger, 9 years old or younger, or 8 years old or younger. The growing conditions for the related plant are not particularly limited, but it is preferable that the environment satisfies, for example, any of the following conditions and is suitable for the growth of the tree species: an average annual temperature of usually 10 to 30°C, preferably 12 to 28°C; and a rainfall of 500 to 3,000 mm, preferably 800 to 2,500 mm.

[0018] [Bark percentage] In this specification, the bark ratio and the trunk ratio are expressed by the following formula: The bark ratio and the trunk ratio have a relationship in which the sum of them is 100%. Bark rate (%) = (bark thickness x 2 / breast height diameter) x 100 Stem percentage (%) = 100 - Bark percentage (%) The breast height diameter and bark thickness can be measured by conventional methods from a cross section of the trunk at breast height.

[0019] In the present invention, a mathematical model is created that correlates differences in DNA sequence with differences in bark ratio (stem ratio), and a predicted value for bark ratio (stem ratio) of a test plant can be obtained from the DNA sequence based on this model. Bark ratio is generally expected to be significantly influenced not only by genetic traits such as variety and lineage, but also by various conditions such as the subsequent growing environment. Therefore, it is difficult to specifically identify genetic markers related to bark ratio from among genetic markers contained in cuttings, such as single nucleotide polymorphisms and microsatellites. By using the predicted value for bark ratio (stem ratio) in the present invention, more appropriate predictions can be made based on conditions such as the seedling environment.

[0020] [Genetic mutation information] The genetic variation information may be of the same species as the genetic variation information of the test plant.

[0021] [Correlation analysis (statistical analysis) and analytical model formula] Examples of statistical analysis methods include the rrBLUP method (Ridge-Regression Best Linear Unbiased Prediction), the GBLUP method (Genomic Best Linear Unbiased Prediction), the GAUSS method (Gaussian kernel regression), the Random Forest method, the RKHS method (Reproducing Kernel Hilbert Space), the Ridge Regression method, the Lasso method, the Elastic Net method, the Bayesian Ridge Regression method, the Bayesian Lasso method, the BayesA method, the BayesB method, the BayesC method, and the like, with the rrBLUP method being more preferred. The programming language may be any language used in statistical analysis such as R, Python, Ruby, Perl, Julia, SAS, SPSS, etc., and it is preferable to use R. Analysis using the rrBLUP method using the programming language R may be performed using an existing script called an existing package. Statistical analysis using the rrBLUP method can be used to obtain an analytical model formula (prediction formula) capable of calculating an estimate of a single trait, bark ratio (stem ratio), by substituting genetic variation information. An example of a method for creating an analytical model is shown below. Based on the bark ratio data and SNP data of a training population of closely related plants (e.g., 99 out of 100 samples), the bark ratio and SNP data are substituted, and the rrBLUP package command (mixsolve) is executed to solve the mixed model, thereby obtaining the best linear unbiased estimator (BLUE), which is an estimate of the fixed effects, and the best linear unbiased prediction (BLUP), which is a predicted value of the random effects, to create a prediction model. An example of the analytical model formula is Equation (1) in Example 1 below. The prediction accuracy of the model formula may be confirmed. Confirmation can be performed using methods such as the leave-one-out method or K-fold cross-validation method.

[0022] [Calculation of predicted bark ratio of test plant] The predicted value of the bark ratio of the test plant can be obtained by applying the genetic variation information of the test plant to the analytical model formula.

[0023] <Application of predicted bark ratio> By obtaining a predicted bark percentage for a test plant, test plants with a small or large predicted bark percentage can be selected, allowing plants predicted to have a desired bark percentage to be selectively used for seedling production, tree production, and afforestation. Test plants with a small predicted bark percentage can be used as wood chips (for papermaking, lumber, etc.). Alternatively, test plants with a large predicted bark percentage can be used for dyes, medicinal ingredient extraction, crafts, fibers, Japanese paper, building materials, etc. [Example]

[0024] The present invention will be described below with reference to examples, but the present invention is not limited to these examples.

[0025] <Example 1> Creation of a prediction formula for predicting the bark ratio of each plantation tree and verification of its accuracy Using 574 eucalyptus varieties as target samples, we created a prediction formula for predicting the bark ratio of plantation trees and verified its accuracy (Figure 1). [Test method] (1) Acquisition of trait data (bark ratio, trunk ratio) The bark at breast height was peeled from 6.0-year-old eucalyptus plantations (Eucalyptus Pellita, E. urophylla, E. brassiana, E. uro-grandis 574 line) and the bark thickness and diameter at breast height were measured. The bark ratio (bark thickness x 2 / diameter at breast height) and trunk ratio (1 - bark ratio) were then calculated (Table 1). The growing environment for the plantations was an average annual temperature of 27°C, rainfall of 2200 mm, and clay to sandy soil.

[0026] [Table 1]

[0027] (2) Obtaining genetic polymorphism data Analytical samples were collected from the leaves of the 574 Eucalyptus accessions used in (1). DNA was extracted from the collected samples using the method described in Silva et al. New Phytologist (2015) 206:1527-1540. SNP data (single nucleotide polymorphism data) were obtained using the Axiom Eucalyptus Genotyping Array (EMBRAPA, Brazilian Agricultural and Livestock Corporation). This resulted in the extraction of approximately 47,000 confirmed polymorphic SNPs with a call rate of over 90% (Figure 2).

[0028] (3) Statistical analysis Prior to statistical analysis, the SNP data were preprocessed. This involved converting the SNP data into numerical values ​​(e.g., T:T = +1, T:A = 0, A:A = -1) and imputing missing data. Missing data was imputed by calculating the overall average value of the missing data and substituting it. Based on the preprocessed SNP data and the bark thickness and bark ratio (stem ratio) data for 574 individuals, the statistical analysis software R ("The R Project for Statistical Computing" https: / / www.r-project.org / ) was used to clarify the effect (marker effect) per SNP on bark thickness and bark ratio (stem ratio) using the rrBLUP package. The SNP analysis model formula using the rrBLUP method is shown below.

[0029]

number

[0030] A prediction model was created and 10-fold cross validation was performed 100 times to confirm the accuracy of the prediction formula. The prediction accuracy was R = 0.411 for bark thickness, R = 0.438 for bark ratio, and R = 0.438 for stem ratio, with slightly higher accuracy for bark ratio and stem ratio (Table 2).

[0031] [Table 2]

[0032] (4) Confirmation of prediction accuracy using test samples Next, the predictive accuracy of each prediction formula was confirmed using test samples of 3.0-year-old trees (128 samples in Table 3) that had SNP data and trait data for bark ratio and stem ratio but were not used when creating the prediction formula.

[0033] [Table 3]

[0034] When checking the prediction accuracy for test samples, differences in prediction accuracy were confirmed, with the prediction accuracy for the bark ratio and trunk ratio (bark thickness = 0.418, bark ratio = 0.461, trunk ratio = 0.461) being higher (Table 4). This reaffirmed that values ​​evaluated as a percentage of the bark and trunk, taking into account differences in DBH, rather than simple bark thickness, can be predicted with higher accuracy.

[0035] [Table 4]

[0036] Example 2: Examination of the number of samples required for a prediction model The 574 samples used in Example 1 (1) were classified by DBH, and were classified into DBH 8 cm or more (528 samples), DBH 10 cm or more (460 samples), DBH 12 cm or more (359 samples), DBH 14 cm or more (279 samples), DBH 16 cm or more (191 samples), and DBH 18 cm or more (110 samples). From these, a prediction formula for stem rate was created and the prediction accuracy was confirmed, and the prediction accuracy was confirmed using 126 test samples that were not used when creating the prediction formula, as in Example 1.

[0037] On the other hand, the same study was conducted except that the number of samples was randomly reduced so that it was the same as the number of samples in the DBH classification above. The study was conducted three times (a decision was made each time as to which samples to reduce), and the average value of the three times was calculated.

[0038] As a result, when checking the accuracy of the prediction model, it was confirmed that the prediction accuracy was higher when the number of samples was reduced only for samples with large DBH (blue) compared to the prediction accuracy when the number of samples was reduced randomly (orange) (Figure 3). A similar trend was observed for bark thickness (not shown).

[0039] Furthermore, in verification using test samples, it was revealed that the prediction accuracy was higher when the number of samples was reduced by DBH (blue) compared to the prediction accuracy when the number of samples was reduced randomly (orange) (Figure 4).

Claims

1. (A): Obtaining genetic variation information of a test plant that is a woody plant; and (B): Obtaining a predicted value of the bark ratio of the test plant based on an analytical model formula obtained by statistically analyzing the correlation between the genetic variation information of at least two closely related plants of the test plant and the bark ratio of each closely related plant from the genetic variation information of the test plant. Including, A method for predicting the bark percentage of a test plant.

2. The method according to claim 1, wherein the closely related plant is a plant having a diameter at breast height of 10 cm or more.

3. 3. The method according to claim 1 or 2, wherein the statistical analysis is performed by a method selected from the group consisting of rrBLUP method, GBLUP method, GAUSS method, Random Forest method, RKHS method, Ridge Regression method, Lasso method, Elastic Net method, Bayesian Ridge Regression method, Bayesian Lasso method, Bayes A method, Bayes B method, and Bayes C method.

4. 3. The method according to claim 1 or 2, wherein the closely related plants include plants selected from plants of the same genus as the test plant but of different varieties or lineages.

5. The method according to claim 1 or 2, wherein the genetic variation information is information about a single nucleotide polymorphism.

6. The method according to claim 5, wherein the number of single nucleotide polymorphisms is 500 or more.

7. The method according to claim 5, wherein the number of single nucleotide polymorphisms is 1,000 to 47,000.

8. 3. The method according to claim 1, wherein the test plant is a Eucalyptus plant.

9. A method for selecting plants for wood chips, comprising obtaining predicted bark percentage values ​​for at least two test plants using the method described in claim 1 or 2, and selecting test plants with low predicted bark percentage values.

10. 10. The method of claim 9, wherein the wood chips are for paper or fuel use.

11. A method for producing seedlings, comprising producing seedlings of plants selected by the selection method according to claim 9.

12. A method for producing wood chips, comprising growing trees from seedlings produced by the production method of claim 11, and harvesting the trees and processing the wood obtained into wood chips.

13. 13. The method of claim 12, wherein the wood chips are for fuel.

14. 13. The method of claim 12, further comprising debarking the wood or wood chips.

15. 15. The method of claim 14, wherein the wood chips are for papermaking.

Images