These experiments used whole-genome arrays from Agilent Technologies. They were constructed of 60-mer oligonucleotides designed to match predicted genes in D. melanogaster genome release 3.1. Thus, these arrays are very similar to those used by Gibson et al. (2004). The major difference is that these new arrays were specially designed to discriminate among alternate transcripts of individual genes and among transcripts of multi-gene families. In total, there are data for approx. 10,000 genes. The platform description is available from the GEO database under accession GPL3809.
Two parental strains were used for experiments, Oregon-R and Russian 2b. In addition, six recombinant inbred (RI) lines were surveyed. However, for Sebida we include only data for the parental strains. cDNA was derived from virgin males and females 3-4 days of age after an amplification step using Agilent protocols. For each of the two parental strains, 8 hybridizations were performed - 4 with male cDNA and 4 with female cDNA. The male and female cDNA were labeled with different fluorescent dyes and hybridized on the same array. Biological and dye-swap replicates were performed. The male/female ratios given in Sebida are averages over all replicates, taken from the authors' supplemental file. The raw data are available from the GEO database under the accession numbers: GSM111773, GSM111774, GSM111775, GSM111776, GSM111777, GSM111778, GSM111779, GSM111780, GSM111781, GSM111782, GSM111783, GSM111784, GSM111785, GSM111786, GSM111787, GSM111788.
P-values were calculated from the raw data using a Bayesian method (BAGEL) developed by Jeffrey Townsend. Because BAGEL requires male/female expression ratios, we paired the results of male and female hybridizations performed on the same array. BAGEL analysis was performed separately for the Oregon-R and Russian 2b strains, as well as for both strains combined.
In cases where there was more then one probe per gene, we used the average over all probes. In order to make the results comparable to other experiments that did not discriminate among alternative transcripts, this was also done for genes with sexually dimorphic expression of alternate transcripts. However, in these cases, we have added a "flag" in Sebida to indicate that different transcripts show expression differences between the sexes. Full information on these can be obtained from the author's supplemental file. Note that the ratios presented in Sebida are straight ratios and not ln or log2 transformations.