Identification of candidate genes associated with phenotypic plasticity through GWAS
Within the LD interval upstream and downstream of the significant SNPs detected by GWAS, the B. napus homologs for seven candidate genes known to participate in the lipid metabolism of Arabidopsis seeds (http://aralip.plantbiology.msu.edu/) were detected: PMT6(BnaA05G0437100ZS), DAO1 (BnaA09G0636200ZS), MYB106(BnaA01G0418000ZS), HAD (BnaC02G0039300ZS), MIPS3(BnaC02G0039400ZS), DGD1 (BnaC05G0484900ZS) and PSD1(BnaC09G0366500ZS).
To investigate the variation in the promoter regions of these genes, the methodology described by Li was followed(Li et al., 2021). In detail, sequences surrounding the genes (PMT6 and DGD1 ) implicated in SOC phenotypic plasticity were retrieved from the B. napuspan-genome information resource (BnPIR; http://cbi.hzau.edu.cn/bnapus/) for the inbred lines ZS11, Zheyou7, Gangan, Shengli, No2127, Westar, Quinta and Tapidor. The sequences of these two genes were also extracted from this website. The colinear sequences among these eight inbred lines were used as queries for a BLAST search against each other to identify structural variations. Transposon elements (TEs) present in the sequences were identified with a TE Library constructed by Extensive de-novo TE Annotator (EDTA) (Ou et al., 2019). In addition, sequence variation within the genes (DAO1, MYB106 and HAD ) was analyzed with polyphen-2 (http://genetics.bwh.harvard.edu/pph2/) and provean (http://provean.jcvi.org/), noting putative loss-of-function alleles for each gene as predicted by at least one of the software platforms (Supplementary Table S16).
To model the genetic effect dynamics along the environmental gradient, SNPs significantly associated with slope were tested for association with SOC within each environment using the mixed-model method in GAPIT. These separate genetic effects were then regressed on the environmental gradients to generate the fitted lines as the genetic effect continua, defined as varied genetic effects with different environmental inputs (Fig. 5D-F).
We used RNA sequencing data from 289 lines (subset of 505 lines) of developing seeds at 20 days after flowering in WH2016 (Tang et al., 2021). The expression levels of the plasticity genes were determined by the abundance of transcripts and compared between two alleles with a Wilcoxon test.