Fig. S1 Workflow for generation of the phenotypic and genotypic data sets for association mapping.

Fig. S2 The distribution of clonal effects (H2) as measured by the fraction of phenotypic variance accounted for by clonal identifiers in an ANOVA with clone as a fixed effect.

Fig. S3 Clonal means for the 292 metabolites were largely uncorrelated, as assessed with Spearman’s rank correlation (ρ), with one another.

Fig. S4 Standardized Gene Ontology (GO) terms for the 1487 out of 2488 EST contigs with a significant BLAST hit to a gene model in Arabidopsis that hit a term nested under molecular function (GO:0003674).

Fig. S5 Pairwise plots of the top four genetic principal components (PCs) derived from a principal components analysis (PCA) on the full 3563 SNP data set.

Fig. S6 Cumulative distribution plots of P-values from single SNP association tests using the (n − k − 1) r2-statistic.

Fig. S7 The distribution of minor allele frequencies (MAFs) for the top 24 SNPs, as measured using the FDR Q-value across 500 randomizations of phenotypic vectors relative to genotypic vectors.

Fig. S8 SNPs associated to at least one metabolite exhibited different minor allele frequencies (MAFs) and magnitudes of genetic differentiation (FST) among populations relative to the entire set of SNPs.

Fig. S9 Distribution of the fraction of clonal effects captured by the adjusted R2 from a linear model relating multiple ancestry-corrected SNPs to ancestry-corrected phenotypes for unknown metabolites.

Fig. S10 Distributions of additive and dominance effect sizes for unknown and known metabolites for genetic associations involving SNPs with three genotypic categories.

Fig. S11 Additive and dominance effect sizes were correlated with one another and with the minor allele frequency (MAF).

Table S1 A summary of known metabolites detected using GC-TOF-MS

Table S2 A summary of unknown metabolites detected using GC-TOF-MS

Table S3 Summary of linkage disequilibrium among SNPs used for association mapping

Table S4 Summary of associations for mannitol as identified using Bayesian mixed linear models

Table S5 Metabolomic phenotype data

Table S6 SNP genotype data

Table S7 Attributes of the genetic loci containing the SNPs used for association mapping

Table S8 Genetic associations detected using a multilocus Bayesian model (BAMD)

Methods S1 Discovery of single nucleotide polymorphisms.

Methods S2 Functional annotation of EST contigs.

Methods S3 Site annotations of SNPs.

Methods S4 Mode of inheritance.

Methods S5 Randomization tests.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

NPH_3976_sm_FigS1-S11-TableS1-S4-MethodsS1-S5.doc1483KSupporting info item
NPH_3976_sm_TableS5-S8.xls5635KSupporting info item