Table 1.
Statistics of sequencing and assembly data.
Figure 1.
Length distribution (bp) of the 38,297 contigs of the global 454 assembly.
Figure 2.
GO enrichment analysis between fiber development stages.
The Gossip package of Blast2GO was used. Over representation of functional classes are presented among contigs with significant DGE (over-expression) at 10 compared to 22 dpa (upper Figure) and at 22 compared to 10 dpa (lower Figure). Significantly enriched GO terms (P<0.05) are highlighted, and the degree of color saturation of each node positively correlates with the enrichment significance of the corresponding GO term. Box color represent significance ranging from white for least significant to dark orange for most significant enrichment, arrows indicate hierarchical relationships.
Table 2.
Most highly abundant transcripts (contigs with >1,000 reads).
Figure 3.
Number of contigs in the different cell wall-related gene categories.
The gene categories definitions refer to the 2nd or 3rd organization levels in Cell Wall Navigator, http://bioweb.ucr.edu/Cellwall/). Number of reads are indicated in square brackets.
Figure 4.
Changes in expression of fiber genes for all possible pair-wise comparisons.
Comparisons are contrasted between: A- fiber development dates (10 vs. 22 dpa), B-genotypes (G. hirsutum and G. barbadense, Gh vs. Gb), and C-sub-genomic origin of reads (A sub-genome vs. D sub-genome). Gene expression was assessed digitally according to the origins of reads within contigs. In each pair-wise comparison bar length is proportional to the number of contigs showing differential over-expression (number of contigs also indicated in parentheses). Symbol ‘*’ indicates a statistical difference (P0.05) for each pair-wise comparison using Fisher’s exact test. For example, within the A panel (all comparisons involving 10 vs. 22 dpa), the comparison Gh10_A vs. Gh22_A (164 vs. 102, significantly different) indicates that within G. hirsutum, there are 164 contigs for which the A-tagged reads are more abundant (≥2-fold) at 10 dpa than at 22 dpa; conversely 102 contigs have more abundant representation of A-tagged reads at 22 dpa than at 10 dpa.
Table 3.
Genes most differentially expressed between the 2 stages of fiber development.
Table 4.
Summary table of SSR mining in the contigs.
Figure 5.
Schematic partitioning of the 39,099 variant positions between intra- and inter-genotypic SNPs.
Variant positions are counted among contigs covered by the 2 genotypes and with at least 6 reads of both. The blue and brown bars symbolize the AT and DT sub-genome co-assembled homoeo-copies. Genotypes are symbolized in plain (G. hirsutum) and dotted (G. barbadense) lines.