Method and system for identifying rice drought-resistant gene based on multispectrum

Abstract

The invention relates to the field of crop information, and discloses a multispectral-based method and system for identifying a rice drought-resistant gene, and the method comprises the following steps: S1, planting a rice genetic material under a drought stress condition and a normal condition, and obtaining multispectral image data; s2, processing the multispectral image data to obtain multispectral phenotypic characters related to the drought resistance of the rice; s3, based on the multispectral phenotypic characters and the genome data after quality control, performing whole genome association analysis through a mixed linear model, and positioning drought-resistant candidate genes; s4, performing eQTL analysis by combining expression quantity data of the drought-resistant candidate gene under drought stress conditions and normal conditions, and verifying a drought-resistant function. The method provided by the invention not only overcomes the defect of lossy measurement in a method for measuring physiological and biochemical indexes related to drought resistance during traditional rice drought-resistant gene identification, but also provides a new and feasible solution for extracting useful information from massive multispectral data.

Description

TECHNICAL FIELD

[0001] The application relates to the field of crop information, in particular to a method and system for identifying drought-resistant genes of rice based on multi-spectrum. BACKGROUND

[0002] Rice is not only one of the three major crops in the world, but also one of the crops with the largest water requirement. However, many major rice-producing areas are facing a shortage of agricultural water resources, which affects the yield of rice. The use of multi-spectrum phenotype acquisition equipment combined with image processing technology can realize the non-destructive and accurate measurement of drought-resistant related indexes, identify drought-resistant genes of rice, and provide new genes for drought-tolerant rice breeding, which is of great significance for the research of drought-tolerant rice breeding.

[0003] The prior art proves that the BADH gene can effectively improve the drought resistance of rice plants under drought stress by determining and analyzing the MDA content, SOD enzyme activity, proline content and leaf 24h water loss rate of the flag leaf after flowering of two BADH gene lines.

[0004] Some people have also studied the effects of drought stress on the activities of antioxidant enzymes and the contents of osmotic adjustment substances in drought-sensitive and drought-tolerant rice varieties during the tillering stage. The results show that spraying 6-BA and BR on the rice leaves during the tillering stage can effectively alleviate the effects of drought stress on physiological characteristics.

[0005] In addition, some studies have also researched the effects of drought training during the tillering stage of rice on the development of panicles and yield formation of rice during the young panicle differentiation stage under high temperature, and explored the physiological mechanisms. Some researchers have conducted a whole-genome association analysis on 21 root traits such as root length, root dry weight and root volume of 529 rice varieties under drought stress, and obtained 143 significant association sites, 80% of which are located within the previously reported QTL range. These sites include genes such as DRO1, WOX11, OsPID, Nal1 and OsJAZ1 that control root drought traits.

[0006] In 1982, researchers proposed using normalized difference vegetation index (NDVI) to monitor the physiology and structure of plants in a large area in the field of spectral imaging. Spectral imaging has increased the ability to obtain data during the crop breeding process, opening up a new field of data acquisition from images. High-throughput phenomics can not only be used for rapid screening of crop varieties with excellent drought tolerance traits, but also can reveal the physiological and molecular response mechanisms of crops to environmental stresses such as drought. By integrating research results from multiple disciplines such as genetics, molecular biology and ecology, a new perspective is provided for crop adaptability research, promoting the development of precision agriculture and sustainable agriculture. SUMMARY

[0007] In order to solve the problem of damaging measurement in the method for measuring drought resistance related physiological and biochemical indexes in the identification of existing drought resistance genes of rice, the application provides a method for identifying drought resistance genes of rice by using multi-spectral phenotype, realizes non-destructive and accurate measurement of drought resistance related indexes by combining with image processing technology, and completes mining of new genes by whole genome association.

[0008] In order to achieve the above-mentioned purpose, the application provides a method for identifying drought resistance genes of rice based on multi-spectrum, and the steps include:

[0009] S1. Planting rice genetic materials under drought stress conditions and normal conditions to obtain multi-spectral image data;

[0010] S2. Processing the multi-spectral image data to obtain multi-spectral phenotype related to drought resistance of rice;

[0011] S3. Based on the multi-spectral phenotype and the quality controlled genomic data, performing whole genome association analysis by mixed linear model to locate drought resistance candidate genes;

[0012] S4. Combined with the expression amount data of the drought resistance candidate genes under drought stress conditions and normal conditions, performing eQTL analysis to verify drought resistance function.

[0013] Preferably, the S1 includes: using 199 kinds of rice genetic materials from the 3K rice core germplasm resource library and 40 kinds of rice genetic materials from the genetic development resource library for pot planting; daily manual measurement of plant height, soil temperature and soil humidity, wherein the plant height is measured by a horizontal reference plate, the soil temperature and humidity are measured twice at a depth of 20 cm, and the average value is taken.

[0014] Preferably, in the S2, the processing step includes: collecting DAT format multi-spectral data of different wave bands, converting the data into images, generating a binary image based on a chlorophyll channel threshold value, extracting an ROI region after median filtering denoising; calculating DN values and derived parameters of each wave band to obtain digital phenotype parameters.

[0015] Preferably, after the multi-spectral image data is processed, the multi-spectral phenotype related to drought resistance of rice is obtained by dimension reduction screening, and the step of dimension reduction screening includes: sequentially performing Min-Max normalization, 1.5IQR outlier rejection, batch repeatability test, independent sample T test, principal component analysis, multi-layer perceptron screening and genetic force analysis on the phenotype parameters.

[0016] Preferably, the step S3 includes:

[0017] Quality control is performed on the genomic data to filter SNP sites with a missing rate of >20% and MAF of <0.05;

[0018] The genotype and phenotype are associated by using a mixed linear model, wherein the population structure matrix and the kinship matrix are random effects, and the soil temperature and humidity are fixed effects;

[0019] The whole genome correlation analysis is performed by the P3D method of the TASSEL5.0 software, and the associated SNP adjacent genes are located in combination with genome annotation.

[0020] Preferably, the S4 comprises:

[0021] The eQTL analysis is performed on the candidate genes, and the p value is corrected by using FDR.

[0022] The gene function is judged by the difference between the eQTL signals under the drought stress condition and the normal condition.

[0023] The drought resistance phenotype verification is performed by using the mutant or knockout material of the candidate gene.

[0024] Preferably, the standard for judging the gene function comprises that only the significant one under the drought stress condition is a drought resistance response gene, and the significant one under both conditions is a basic regulation gene.

[0025] The application further provides a system for identifying drought resistance genes of rice based on multi-spectrum, which is used for realizing the above method and comprises a collection module, a processing module, a positioning module and a verification module.

[0026] The collection module is used for planting the genetic material of rice under the drought stress condition and the normal condition, and obtaining multi-spectrum image data.

[0027] The processing module is used for processing the multi-spectrum image data, and obtaining multi-spectrum phenotype traits related to the drought resistance of rice.

[0028] The positioning module is used for performing whole genome correlation analysis by using a mixed linear model based on the multi-spectrum phenotype traits and the quality-controlled genome data, and positioning drought resistance candidate genes.

[0029] The verification module is used for performing eQTL analysis in combination with the expression data of the drought resistance candidate genes under the drought stress condition and the normal condition, and verifying the drought resistance function.

[0030] Compared with the prior art, the application has the following beneficial effects:

[0031] The application not only overcomes the damage measurement in the method for measuring drought resistance related physiological and biochemical indexes in traditional rice drought resistance gene identification, but also provides a new and feasible solution for extracting useful information from massive multispectral data, and realizes nondestructive, accurate and rapid measurement of rice drought resistance related indexes. Based on the method for identifying rice drought resistance genes based on multispectral, the feasibility of multispectral technology, image processing technology and whole genome association analysis technology in rice drought resistance gene identification is verified, and a bridge between multispectral imaging technology and drought resistance gene identification is established, thereby providing a new method for rice drought resistance gene identification. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the present application, the following briefly introduces the drawings needed in the embodiments. Obviously, the drawings described in the following embodiments are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0033] Figure 1 The technical solution flowchart of the present application. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0035] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail with reference to the drawings and specific embodiments.

[0036] Embodiment one:

[0037] The present embodiment provides a method for identifying rice drought resistance genes based on multispectral, and the steps include:

[0038] S1. Planting rice genetic materials under drought stress conditions and normal conditions to obtain multispectral image data.

[0039] Pot planting of 199 rice genetic materials from the 3K rice core germplasm resource library and 40 rice genetic materials from the core germplasm resource library of the Institute of Genetics and Developmental Biology, Chinese Academy of Sciences under drought stress conditions and normal conditions; using artificial measurement of plant height, soil temperature, and soil humidity as artificial measurement values; fitting plant height with digital plant height extracted based on multispectral data to ensure data accuracy; soil temperature and soil humidity ensure the growth state of rice and the effectiveness of drought treatment. Each variety is planted in a plug tray after germination and seedling growth in the same environment, and transplanted at 24 days. Six plants with consistent growth and size are selected for transplanting, with three plants under drought conditions and three plants under normal conditions.

[0040] S2. Process the multispectral image data to obtain multispectral phenotypic traits related to drought resistance of rice.

[0041] Multispectral data is collected for 14 consecutive days after rice transplanting. The collected results are DAT format files, which are converted to image format and processed to extract digital phenotype parameters. After data normalization, phenotypic trait dimensionality reduction processing and key trait screening are performed, including outlier screening, repeatability testing, independent sample T-test, principal component analysis, multilayer perceptron screening, and genetic force analysis, to further screen multispectral phenotypic traits with high genetic force and high correlation with rice drought resistance.

[0042] The artificial measurement of plant height in step S1 is fitted with the digital plant height extracted based on multispectral data in step S2, and the fitting result has a correlation of 99% and an R² (value between 0 and 1, where R²=1 is a perfect model fitting data and R²=0 is a model that cannot explain data variation) of 98%, ensuring data accuracy and correlation. Artificial measurement of soil temperature and soil humidity can clearly observe soil changes to ensure the growth state of rice and the effectiveness of drought treatment.

[0043] More specifically, step S2 uses the phenotyping device Plant Explorer to collect multispectral data of the growth conditions of potted rice of different varieties under drought stress conditions and normal conditions, which can obtain DAT format data of Blue (475 nm), Green (550 nm), Red (640 nm), Chlorophyll (730 nm), Anthocyanin (540 nm), Red-Edge (710 nm), Nir-Red (770 nm), F0 (initial fluorescence), Fm (maximum fluorescence yield), etc. After reformatting the imaging data, image segmentation and image masking are performed to extract digital phenotype parameters and perform data analysis.

[0044] Image processing and data analysis use python3.12 to reorganize the obtained data into multispectral images in different wave bands, take the Chlorophyll (730nm) channel sensitive to plants for threshold segmentation, set the threshold value to 1200, set the pixels greater than 1200 to white and other areas to black, to obtain a binary image, and use median filtering and small area denoising method to remove noise.

[0045] Extract digital phenotype parameters, mask all multispectral images using the obtained binary image, and obtain ROI multispectral images in each wave band; calculate the DN value in different wavelengths using the ROI multispectral image, and use the ROI image in different channels to perform arithmetic operation to obtain derived parameters. According to the above digital phenotype traits, a total of 352 parameters can be obtained, and after data cleaning, 348 phenotype parameters are retained.

[0046] All phenotype parameters are normalized using Min-Max normalization:

[0047] ;

[0048] Wherein, X max , X min , X represent the maximum value, minimum value and value to be normalized in the group of phenotype data.

[0049] Phenotype traits need to be processed by dimension reduction and key trait screening to obtain suitable phenotype parameters, including outlier screening, repeatability test, independent sample T test, principal component analysis, multilayer perceptron screening and heritability analysis.

[0050] Outlier screening uses the 1.5IQR standard. If the lower quartile of a group of data is Q1, the upper quartile is Q3, and the interquartile range is IQR, most data will be within the range of [Q1 - 1.5IQR, Q3 + 1.5*IQR] under normal circumstances, and data outside this range is considered abnormal and can be removed.

[0051] Repeatability screening can ensure that the trait has stability between different experimental batches. Using pandas and scipy.stats libraries of Python, the Pearson correlation coefficient between each trait and different batches is calculated, and all traits with a correlation coefficient ≥0.8 between batches are retained for subsequent analysis.

[0052] Independent sample T test can eliminate phenotype traits with no significant difference between normal and drought conditions. Using scipy.stats library, the confidence interval is set to 95%, and through independent sample T test, 90% of the phenotype data can be analyzed further.

[0053] PCA can reduce the dimension of data while retaining the main variation. The data is standardized using Python's StandardScaler, and the number of principal components (explaining 95% of the variance) is specified. The variance contribution rate and cumulative variance contribution rate of the principal components are analyzed, and the principal component scores are retained for subsequent analysis.

[0054] MLP is used to select important multispectral phenotypic traits. Python's scikit-learn library is used for MLP, and the experiment is repeated five times. The importance of the traits is calculated, and the phenotypic traits with an average importance of more than 50% are retained for subsequent analysis.

[0055] Genetic analysis and reproducibility analysis use the method of ANOVA. ANOVA estimates the genetic force of the trait by decomposing the total variation of the trait into genetic variation and environmental variation. Through ANOVA, the contribution of genetic factors to the variation of the trait can be quantified, and the genetic basis and heritability of the trait can be evaluated. Genetic force is the proportion of variation caused by genetic factors in the variation of the trait. The formula is:

[0056] ;

[0057] ;

[0058] where h 2 is the narrow-sense heritability, H 2 is the broad-sense heritability, is the genetic variance, is the environmental variance, is the error variance.

[0059] In the reproducibility analysis, ANOVA decomposes the total variation of the trait into environmental variation and measurement error variation. Through ANOVA, the influence of environmental factors and measurement error on the variation of the trait can be quantified, and the reproducibility and stability of the trait can be evaluated. The formula is:

[0060] ;

[0061] where, is the sum of genetic variance, environmental variance, and error variance.

[0062] S3. Based on the multispectral phenotypic traits and the quality-controlled genomic data, whole-genome association analysis is performed through mixed linear model to locate drought-resistant candidate genes.

[0063] The mixed linear model is used to perform whole-genome association analysis on the screened multispectral phenotypic traits and the quality-controlled genomic data. Combined with rice genome annotation, significant SNP adjacent genes and hot genes are found as candidate genes for rice drought resistance function.

[0064] Specifically, the S3 genomic data quality control needs to filter SNP sites with high deletion rate (sites with deletion rate greater than 20% will be removed), high-quality SNPs (MAF≥0.05) markers, ensure that the markers have sufficient population representation, and reduce the noise caused by low-frequency variations.

[0065] Genotype and phenotype association is performed by using a mixed linear model (MLM), potential confounding factors are controlled by introducing a population structure matrix (Q) and a kinship matrix (K), known environmental factors that can affect the phenotype, such as soil temperature and soil humidity, are taken as fixed effects in the model to further reduce environmental noise; the P3D method without compression in TASSEL5.0 software is used for whole genome association analysis.

[0066] S4. eQTL analysis is performed in combination with the expression amount data of the drought-resistant candidate gene under drought stress conditions and normal conditions to verify the drought-resistant function.

[0067] The drought-resistant function of the candidate gene associated with the multi-spectral phenotype trait is verified, eQTL analysis is performed in combination with the expression amount data of the candidate gene under normal watering and drought stress, and the drought-resistant function of the candidate gene is verified by using the mutant or transgenic material of the candidate gene. Specifically, FDR corrected p value is used for multiple test correction, and FDR < 0.05 is generally considered significant.

[0068] Compare the eQTL signals under drought and normal conditions: if the eQTL is only significant under drought conditions, it means that the QTL can respond to drought stress, regulate gene expression, and complete the drought resistance reaction; if the eQTL is significant under both conditions, it means that the QTL is involved in the basic reaction and participates in the expression regulation. Select drought-resistant candidate genes based on eQTL analysis; use the mutant or gene knockout material of the candidate gene to verify the drought-resistant phenotype and confirm the drought-resistant function of the candidate gene. The technical solution process of this embodiment is shown in Figure 1 .

[0069] Embodiment Two

[0070] The embodiment also provides a system for identifying drought-resistant genes of rice based on multi-spectrum, which comprises a collection module, a processing module, a positioning module, and a verification module; the collection module is used for planting rice genetic materials under drought stress conditions and normal conditions to obtain multi-spectral image data; the processing module is used for processing the multi-spectral image data to obtain multi-spectral phenotype traits related to the drought resistance of rice; the positioning module is used for locating drought-resistant candidate genes by performing whole genome association analysis on the multi-spectral phenotype traits and the quality-controlled genomic data through a mixed linear model; and the verification module is used for performing eQTL analysis in combination with the expression amount data of the drought-resistant candidate gene under drought stress conditions and normal conditions to verify the drought-resistant function.

[0071] The above-described embodiments are merely intended to describe the preferred modes of the present application, and are not intended to limit the scope of the present application. Various modifications and improvements to the present application made by those skilled in the art, without departing from the design spirit of the present application, shall fall within the scope of the present application as defined by the claims.

Claims

1. A method for identifying drought-resistant genes in rice based on multispectral identification, characterized by the steps of The method comprises the following steps: S1. Planting rice genetic materials under drought stress conditions and normal conditions to obtain multispectral image data; S2. Processing the multispectral image data to obtain multispectral phenotypic traits related to drought resistance of rice; S3. Based on the multispectral phenotypic traits and the quality-controlled genomic data, performing whole genome association analysis by mixed linear model to locate drought resistance candidate genes; S4. Combined with the expression data of the drought resistance candidate genes under drought stress conditions and normal conditions, performing eQTL analysis to verify the drought resistance function.

2. The method for identifying drought-resistant genes in rice based on multispectral identification according to claim 1, characterized in that, The S1 comprises: using 199 kinds of rice genetic materials from the 3K rice core germplasm resource library and 40 kinds of rice genetic materials from the genetic development library for pot planting; daily manual measurement of plant height, soil temperature and soil humidity, wherein the plant height is measured by a horizontal reference plate, and the soil temperature and humidity are measured by double measurement at a depth of 20 cm to obtain the average value.

3. The method for identifying drought-resistant genes in rice based on multispectral identification according to claim 1, characterized in that, In the S2, the processing steps include: collecting DAT format multispectral data of different wave bands, converting the data into images, generating binary images based on chlorophyll channel threshold segmentation, extracting ROI region after median filter denoising; calculating the DN value of each wave band and the derived parameter to obtain digital phenotype parameters.

4. The method for identifying drought-resistant genes in rice based on multispectral identification according to claim 3, characterized in that, After the multispectral image data is processed, the multispectral phenotypic traits related to drought resistance of rice are obtained by dimension reduction screening, and the steps of dimension reduction screening include: sequentially performing Min-Max normalization, 1.5IQR outlier rejection, batch repeatability test, independent sample T test, principal component analysis, multilayer perceptron screening and genetic force analysis on the phenotype parameters.

5. The method for identifying drought-resistant genes in rice based on multispectral identification according to claim 1, characterized in that, The steps of S3 include: Quality control of genomic data, filtering SNP sites with a missing rate of >20% and MAF of <0.05; Using mixed linear model to correlate genotype and phenotype, wherein the population structure matrix and the kinship matrix are used as random effects, and the soil temperature and humidity are used as fixed effects; Performing whole genome association analysis by P3D method of TASSEL5.0 software, and locating the genes adjacent to the associated SNPs combined with genome annotation.

6. The method for identifying drought-resistant genes in rice based on multispectral identification according to claim 1, characterized in that, The S4 includes: Performing eQTL analysis on the candidate genes, and using FDR corrected p value; Determine the gene function by the difference of eQTL signal between drought stress conditions and normal conditions; Using the mutant or knockout material of the candidate gene to verify the drought resistance phenotype.

7. The method for identifying drought-resistant genes in rice based on multispectral identification according to claim 6, characterized in that, The standard for determining the gene function includes: only significant in drought stress conditions is drought resistance response gene, and significant in both conditions is basic regulation gene.

8. A system for identifying drought resistance genes in rice based on multispectral, the system being used to implement the method according to any one of claims 1 to 7, characterized in that, The method comprises the following steps: A collection module, a processing module, a positioning module and a verification module; The collection module is used for planting rice genetic materials under drought stress conditions and normal conditions to obtain multispectral image data; The processing module is used for processing the multispectral image data to obtain multispectral phenotypic traits related to drought resistance of rice; The positioning module is used for performing whole genome association analysis by mixed linear model based on the multispectral phenotypic traits and the quality-controlled genomic data to locate drought resistance candidate genes; The verification module is used for carrying out eQTL analysis in combination with the expression amount data of the drought-resistant candidate gene under drought stress condition and normal condition, and verifying the drought-resistant function. The verification module is used for carrying out eQTL analysis in combination with the expression amount data of the drought-resistant candidate gene under drought stress condition and normal condition, and verifying the drought-resistant function.

Images