Table 1.
Cellulose quantities by breeding program and subpopulation in barley culms.
Table 2.
Cellulose quantities by subpopulation defined by STRUCTURE in barley culms.
Table 3.
Regions of the barley genome identified by GWAS as having significant associations with cellulose content.
Fig 1.
Genotypic data and population structure analysis.
(A) Principal coordinates analysis of all lines phenotyped, colour coded by breeding program. (B) STRUCTURE bar plot for K = 6 based on bOPA 1&2 genotyping data for spring barley lines ordered by breeding programs, but colour coded by K value. Please note colours in A. are independent to those in B. Breeding program 1 = Washington (WA), 2 = Montana (MT), 3 = 2- row North Dakota (N2), 4 = Utah (UT), 5 = 6-row North Dakota (N6), and 6 = Minnesota (MN). Colours represent subpopulation defined by shared genetic ancestry. Q value represents proportion of ancestry to a given subpopulation. (C) Output from Structure Harvester showing K as calculated based on ΔK method, in this case K = 6. L(K) represents the likelihood distribution of K, and L”(K) represents the second order rate of change from L(K).
Fig 2.
Phenotypic data used to carry out a genome wide association study (GWAS).
(A) Mean culm cellulose content for 2-row and 6 row spring barley accessions by breeding program. (B) Mean culm cellulose content for all lines to illustrate the distribution of this trait in the barley CAP programs included in this analysis.
Fig 3.
Manhattan plots of culm cellulose content (mg cellulose / mg dry weight) genome wide association scans (GWAS) using the Kinship relationship model.
The-log10 (p-values) from a genome-wide scan are plotted against the position on each of the 7 barley chromosomes. Manhattan plots are displayed for those populations which had associations that passed the significance threshold set by FDR p > 0.05 (-log10P >3.0), i.e. populations 3 and 5. Manhattan plots of the null models are provided for comparison. (A) Population 3 Kinship model. (B) Population 5 Kinship model. (C) Population 3 Null model. (D) Population 5 Null model.
Fig 4.
Co-expression of two groups of genes (HvCesA9, HvCobra1, HvCslF6, and HvChitinase, HvGT1) identified by GWAS as putatively linked to culm cellulose content with HvCesA genes known to be involved in primary and secondary cell wall development.
Transcript abundance across a range of tissues shown in fragment per kilobase of exon per million fragments mapped (FPKM) for group 1, primary cell wall including HvCesA1, HvCesA2, and HvCesA6 (A) for reference, and group 2, secondary cell wall including HvCesA4, HvCesA7 and HvCesA8 for reference (B). Abbreviations for tissues/ developmental stages as follows; EMB = Embryo tissues (germinating), ROO = Root (10cm seedlings), LEA = Shoot (10cm seedlings), INF1 = Inflorescence (0.5cm), INF2 = Inflorescence (1–1.5cm), NOD = Tillers (3rd internode), CAR5 = Grain (5 Days Post Anthesis—DPA), CAR15 = Grain (15 DPA), ETI = Etiolated (10 day seedlings), LEM = Lemma (6 weeks Days After Planting—DAP), LOD = Lodicule (42 DAP), PAL = Palea (42 DAP), EPI = Epidermis (28 DAP), RAC = Rachis (35 DAP), ROO2 = Root (28 DAP seedling), SEN = Senescing leaf (63 DAP).