| With the advent of the big data era,the study on Genome-wide association analysis(GWAS)has become one of the research focuses of bioinformatics,and has been widely applied in many fields such as medicine,agriculture,forestry,animal husbandry,etc.The purpose of this study is to find the potential relationship between phenotypic traits and genotypic data in the whole genome,so as to select gene loci affecting traits and detect the genetic mechanism of biological growth.Although GWAS has a broad application space,most studies only focus on the genetic factors of biomass at a single time point or limited individual time points.In the whole growth cycle,the development of the dynamic traits for organism is related to the time factor,so genetic factors not only affect biomass at one time point,but also govern the whole process of growth.At present,GWAS on dynamic traits is being carried out and gradually developed,but most of the current studies have strict restrictions on the application conditions of the models.Therefore,this dissertation focuses on GWAS for dynamic traits of trees and on this basis the intelligent optimization system for seed orchard layout is studied and developed.The main research contents and conclusions are as follows:1.Based on earliness degree(Earliness-index,E-index),a statistical model of GWAS is established.The method of spline interpolation is adopted to fit the growth process of each genotype,and the values of earliness degree are caculated as an index parameter to distinguish different growth curves.A model on Genome-wide association analysis is constructed and a series of hypothesis tests are used to verify the genetic effects of genetic markers on traits.As a result,the significant SNPs related to the growth process are screened out by permutation test.In order to verify the validity and practicability of our model,this method is applied on the data about the volumes of Populus deltoides F1 population during 24 years and 156362 SNP markers by gene sequencing technologies.2.Based on E-index metric,an improved statistical method of GWAS on E-index vector(Earliness-index Vector,E-indexV)is presented.For the phases of the development of trees,the metho d of optimal breakpoints for growth curve of each genotype is established,and it can divide the life cycle into two phases with the maximum diversity of growth trend for each other.For the same poplar data,significant SNPs affecting the development of volume traits can be detected and be located to specific phase by E-indexV method effectively.The results showes that testcross markers have stronger correlationship of growth phases than intercross markers.3.Genomic selection method is used to predict tree volume and to verify the effectiveness of E-index method.GBLUP(genomic best linear unbiased prediction)method is adopted to construct a linear regression model about SNPs affecting growth of volume and breeding value to predict the volume value of trees at any time point.Furthermore,the effectiveness of the screened SNPs by three GWAS methods(E-index,E-indexV,functional mapping)is evaluated by cross validation method.The experimental results show that genomic selection model based on SNPs screened by E-index and E-indexV methods are much more accurate than other methods.4.An intelligent optimization algorithm is adopted to solve the problem of seed orchard layout based on significant SNPs obtained by E-index method to realize seed orchard optimization deployment.An improved cuckoo algorithm based on long integer is proposed and the performance of the algorithm is evaluated in terms of operating speed and result accuracy.Compared with genetic algorithm,the improved cuckoo algorithm has better performance.The intelligent optimization system of seed orchard is further developed.The optimal arrangement could be automatically obtained by the intelligent system after entering the relevant parameters,which provids a powerful intelligent technical support. |
PDF Full Text Request