Genome-wide Transcript Abundance Analysis

As detailed in [2] (link), whole genome variation in transcript abundance was assessed for both young males and females of each of 40 highly inbred lines using Affymetrix Drosophila 2.0 arrays. RNA was extracted in two independent pools of 15 flies/sex/line (40 lines×2 sexes×2 replicates = 160 samples). The raw array data was normalized using a median standardization. After normalization, an analysis of variance was used to partition variation in expression between sexes, among lines, and the sex×line interaction for each expressed transcript. At a false discovery rate of 0.001, the line term was significant for 10,096 of the expressed transcripts. A regression model identified 414 transcripts among these 10,096 that were significantly associated with the competitive fitness phenotype. The residuals from the regression model were used to compute the genetic correlations for MMC in Figure 5A. Tissue-specific expression data for each of the genes in Module 9 was obtained from [31] (link). The values shown in Figure 5C report the tissue-specific expression of each gene as a fraction its expression across all eleven tissues. The doublesex motif whose logo is shown in Figure 5D represents one of 62 Drosophila melanogaster transcription factors whose position-weight matrices were downloaded from http://www.bioinf.manchester.ac.uk/bergman/data/motifs/. The 5′ UTR of each gene in Module 9 was scored for the presence of all 62 motifs; our protocols for calling a motif present and for assessing enrichment are as described in [2] (link).