Fig. 2: Drivers of transcript abundance variation.

a, Examples of alternative 5′ splice site and intron retention variation among pan-transcriptome genotypes. Top: a 1 nt change (G to T) in seven genotypes caused the selection of an upstream splicing site in chr2H11235. Bottom: splicing was abolished in 11 genotypes in chr3H26163 due to a G to T change. b, Heatmap (PAV) of genes with zero transcript abundance in at least one genotype clustered according to similarity. Grey represents zero detected expression. c, Correlations between gene copy number and transcript abundance for 98 multiple-copy genes across five tissues. The boxplot whiskers show minimum and maximum values, the upper bound of the box represents the 75th percentile, the lower bound of the box represents the 25th percentile and the centerline represents the median. d, Location, magnitude and direction of DEGs across the 141 Mb inversion on chromosome 7H. Statistics were performed using the limma-voom R package with multiple comparison adjustments using the BH procedure. e, Heatmap with Pearson correlations between percent identities in TFBSs in upstream 2 kb regions and percent coherence (within ±30%) in expression values computed for sets of ~3,000 genes showing increasing TPM CV. f, For each set of ~3,000 genes, the percent content of core, shell, cloud and DEGs is reported. DEGs, differentially expressed genes; SS, splice site.