Fig. 5: Joint eQTLs of the pan-gene atlas and GWAS, TWAS, and SMR analysis of 34 field agronomic traits and 8 Bgt isolate infection phenotypes in wheat.

a Schematic diagram for the prediction of candidate gene. The GWAS QTL candidate regions include signals from eQTG and intergenic-eQTL. Candidate genes are predicted using cis-eQTLs through the TWAS, and both cis- and trans-eQTLs through SMR. b Venn diagram of the distribution of candidate genes predicted by the GWAS, the TWAS, SMR and the Spearman correlation coefficient (SCC) between expression levels and phenotypic values. c The number of candidate genes classified by the pan-gene atlas. d Three eQTL regulatory patterns assist in the identification of candidate genes. R1 represents candidate genes regulated by both gene and intergenic regions, R2 represents candidate genes regulated only by the gene region, and R3 represents candidate genes regulated only by the intergenic region. e Types and numbers of regulatory patterns for the candidate genes. f Verification of candidate genes using the Kenong 9204 mutant library. The phenotypes of the mutants and wild-type plants for the predicted candidate genes were analyzed using a two-tailed Student’s t test, with the red line representing the ttest distribution for the candidate genes. The blue–green line represents the t test p value distribution for the randomly selected genes. The same number of random genes were randomly selected as the candidate genes. This process was repeated 10 times. g Proportion of candidate genes validated by the Kenong 9204 mutant library compared with the average proportion of validated genes from ten random sets. h Representative candidate genes verified using the Kenong 9204 mutant library. The candidate genes were predicted to be associated with multiple agronomic traits and were validated for relevant traits (TKW, GW, GN, SPN, and GA) in the mutant library. The x-axis shows WT (wild-type) lines and the corresponding mutant lines of candidate genes, with n indicating the sample size. WT lines were used as the control, and the mutant lines of each candidate gene were compared to the WT line using a two-tailed Student’s t test. Red points represent premature termination (high effect) and yellow points represent non-synonymous mutations (moderate effect) in the candidate gene. The blue–green points represent wild type. The box depicts the median and interquartile range (IQR). The whiskers extend to the most extreme data points within 1.5 × IQR from the quartiles. Source data are provided as a Source Data file.