| Maize has integrates multiple functions such as grain gathering,feeding,and energy application,which has been widely cultivated around the world,and extended to 57 ° north latitude of cold regions.Maize originated in the tropics and is a typical C4 plant.Low temperatures can affect many processes of corn growth.As the largest maize production area in China,Heilongjiang Province suffers from various degrees of chilling threaten in the spring and severely affects the quality and yield of corn.Low-temperature stress has become one of the major abiotic stressors of maize growth and development in Northeast China.Faced with global climate change and the demand of national food security,it is an effective way to improve the ability of maize to withstand risks through evaluating maize germplasm,excavating genes,developing molecular markers,creating maize germplasm for resistant chilling.With the development of high-throughput sequencing technologies,especially the rapid application of single nucleotide polymorphism(SNP)genotypes and simplified genome sequencing technology(GBS),a large number of individuals can be quickly and economically sequenced.Based on linkage disequilibrium(LD)and corn-associated population information rich in genetic background,genome-wide association analysis(GWAS)and genomic selection(GS)technologies can be highly effective applied in searching for genetic factors associated with complex traits and marker-assisted breeding.This study was based on the associative population composed of 282 core inbred lines from the United States and evaluated the cold tolerance of grain traits,germination traits,and seedling traits under low-temperature stress respectively.The genotype data from the GBS and 55 k chipsets and their integration were used,10 traits including germination and seedling traits with cold(frozen)conditions and control temperature conditions,hundred kernel weight,were used as phenotypic traits,and GWAS and GS analysis were carried out,and aims to(1)evaluate and select cold(frozen)tolerance lines in maize germplasm resources,(2)use the genome-wide association(GWAS)method to reveal the mechanism of maize genetic resistance to low temperature stress,screen cold(frozen)tolerance molecular markers,and make a cross comparison of this GWAS study results and other QTL results,(3)identify and analyze related genes,excavate the candidate genes for future gene clone verification,(4)through the technology of genome prediction,evaluate the influence of different prediction methods,the density of different genotype on prediction accuracy,(5)compare the effects of different GWAS results on GS prediction accuracy,and identify the main factors affecting the selection of genomes for resistance to low-temperature traits,provide theoretical support and technical reference for the application of genome prediction.The main results are as follows:1.Take 8 ? for 21 days as low temperature stress condition,we identified 14 inbred lines with germination rate of 100%(B37,A632,B57,C49 A,CI91B,CM37,Co255,EP1,NC310,Oh43 E,R109B,R4,T232 and T234),and 48 inbred lines with germination rate higher than 85%.2.The correlation analysis showed that different parts of the traits were not completely independent,there is still a strong correlation between some traits.The hundred kernel weight was significantly correlated with the total absorb water rate and development of seedling.There was a significant positive correlation between the relative height of seedling under chilling and relative development of seedling under recover,relative height of seedling under recover and relative development of seedling under chilling;grade of frost damage was significantly negative correlation with relative development of seedling under chilling and total absorb water rate;there was a significant correlation between relative germination rate,relative germination days,and relative germination index,and the Pearson correlation coefficient absolute values of phenotype and genetic correlation were higher than 70% and 80% respectively.The absolute value of Pearson correlation of Relative germination index and the relative germination rate of phenotypic and genetic correlation reached 90% and 97% respectively,and we found more pleiotropic SNPs in these two traits;the same traits with two kinds of treatments were significant correlation,relative growth rate and seedling height between cold continuous and cold recovery were significant correlation,and the correlation coefficient reached 0.50 above.3.Compare the relationship between the principal component of individual genotype and 10 traits related to low temperature stress,the result showed that the hundred kernel weight,relative germination index,and grade frost damage were significantly negatively correlated with the genetic composition of the germplasm.The total absorb water was significantly positively correlated with the genetic composition of germplasm.Tropical and subtropical germplasm showed lower values in relative germination rate and relative germination index,this may be due to the fact that this germplasm did not undergo domestication of germination under low temperature,so they were sensitive to low temperature.4.In this study,the estimated breeding value was used as phenotypic data,55 k chip data and GBS data were integrated and used as genotype data,and the genome-wide association analysis was performed by FarmCPU software.A total of 58 significant SNPs associated with low temperature traits were identified.There were 7 markers significantly associated with hundred kernel weight,25 markers significantly associated with traits of germination,20 markers significantly associated with cold tolerance traits of seedlings,and 6 markers significantly associated with cold tolerance traits of seedling,among this markers,ss196436428 both had a high association with relative germination days and relative germination index;14 significant SNPs were coincident with previous markers in the study of maize cold resistance.A total of 191 neighboring SNPs with LD values greater than or equal to 0.8 were detected in the range of-375 kb to 357.05 kb from the significant loci,and the 32 neighboring markers were identical to their corresponding significant SNPs.5.In this study,there were 43 genes had been identified by significant associated SNPs and their corresponding adjacent strong LD loci,of which 14 SNPs that overlapped with previous studies were mapped to 10 related genes.By gathering GO analysis,homologous gene analysis,Arabidopsis homologous gene validation,and previous research results revealed that a total of 13 genes themselves or their homologous genes function involved in the stimuli associated with the cold stimulus-response and Physiological function,including the gene GRMZM5G898867 associated with the HKW trait,the genes GRMZM2G028413,GRMZM2G065829(mybr35),GRMZM2G389768,GRMZM2G057186,GRMZM2G012148,GRMZM2G178486 and GRMZM5G806387,GRMZM2G073535,the genes GRMZM2G125669(aox2-alternative oxidase2),GRMZM2G029258,GRMZM2G164308,GRMZM2G060057 associated with seedling stage traits.The partial genes can be used as an important candidate genes to verify by gene cloning in the future.6.We carried out 5 fold cross-validation for 2 kinds of phenotypic value and 10 traits using 7 kinds of methods with on three kinds of genotypes density.The prediction results showed that the rrBLUP and RKHS methods generally have higher prediction accuracy in the prediction of chill tolerance traits;Variance analysis showed that the influence of genotype density on prediction accuracy was not significant,and the prediction results of low density chip marker were better than other high-density genome data.Comparison with 5 fold cross-validation and leave-one-out cross-validation,it was found that prediction result of the 5-fold cross validation was better than the leave-one-out cross-validation for observing value,but the prediction results are converse for the estimated breeding value which plays leading role by additive effect,and indicated that the heritability has a more significant effect on the leave-one-out cross-validation.7.In this study,we performed 5 fold cross-validation with the GWAS-associated loci as covariate using GBLUP model,based on observe value as phenotype.The results showed that in a small amount of QTNs,the prediction accuracy was significantly improved,but with the increase in the number of QTN,the prediction accuracy has a downward trend,indicating that too many false positive results affect the accuracy of genome selection. |
PDF Full Text Request