Redundancy Analysis (RDA)

To study the relative effect sizes and their statistical significance, we carried out multivariate regression using Redundancy Analysis (RDA). The analysis confirmed that the genotype and phenotype (resistant versus susceptible) effects were highly statistically significant (P = 0.001 and P = 0.004), describing 35 % and 9 % of the overall variation (Table 2). The effect of the inoculation treatment alone was not significant (P = 0.238), but the combined effects of genotype-by-experiment and phenotype-by-experiment were significant (P = 0.001 and P = 0.048, respectively), again suggesting genotype-specific response profiles. Accordingly, the RDA plots displayed clear separation when using genotype and phenotype as a covariate but not with inoculation treatment (Figures 2a, b and c). Venn diagrams of the genes contributing to the separation in the RDA analyses demonstrate that while 87 genes contribute significantly to separation according to genotype, only 7 genes contribute directly to the phenotypic variation and 109 genes to the joint effect of phenotype-by-experiment (Figures 2 d and e).