Genomic sequences identified as transcription start sites (TSS); the 5' ends of cDNAs were identified using 5'-end SAGE.
5'-ends of capped transcripts were isolated by 5'-SAGE. Briefly, a biotinylated adapter to 5'-phosphate ends exposed after cleavage of the 5'-cap by tobacco acid pyrophosphatase (TAP). Double-stranded cDNAs were then cleaved with MmeI to generate short 25-27nt fragments, and biotinylated adapters (along with adjacent 5'end sequence) are purified. These SAGE fragments were then used for Illumina library construction (single end layout).
Reads were mapped to Dmel genome annotation R5.3 using a parallel version of BLAT. Only uniquely mapped reads were considered (allowing for up to 3 mismatches). Associated TSS profiles represent mapped 5'-ends.