Fine Analysis Of GWAS Data Of Rice Drought-Resistance Traits And Construction Of Near-Isogenic Lines

Rice(Oryza sativa)is one of the world’s staple crops,which feeds nearly half of the world’s population.Its yield was severely affected by drought.With the decrease of planting areas in China attributed to adjusting the agricultural production structure and optimum system,improving drought resistance of rice has important significance for yield stability under drought disasters cause by global warming.Based on Zilong Guo’s GWAS study in which a large number of drought-resistance loci were identified,we have initiated near-isogenic line construction for 34 loci.At the same time,four important image traits related to drought response or resistance were fine analyzed by combining population re-sequencing data and RNA-seq data to find the causal genes which may affect those phenotypes.First,we performed variant calling analysis to get the variant information in 636 rice varieties including Kongyu131.Based on the lead SNP genotype in GWAS results,we selected extreme-phenotype varieties as hybrid parents.Finally we designed 37 hybrid combinations and we developed molecular markers to construct near-isogenic lines in order to improve rice drought resistance.Meanwhile,we carried out intensive bioinformatics analysis for the four important drought stress image traits.169 candidate genes were identified with the presence of significant SNPs directly located within the gene according to the previous GWAS results.Then,we used RNA-seq data,which was generated by other colleagues in our group,to identify 66 differentially expressed genes between the two drought related alleles.We also predicted the effects of variations on genes to find out which variation might result in a high impact on the protein structure.We found that seven candidate genes contain high impact In Del variation and 21 candidate genes contain high impact SNP variation.In conclusion,we identified many possible causal genes for the drought-related image traits through bioinformatics analysis and generated backcross lines for construction of NILs which are useful for further studies.