In the GDP database of Exelixis sequences, for d03281 I found two 3' flanks and three 5' flanks (it was common to find multiple independent sequence runs for Exelixis lines). The two 3' flanks can be made into a consensus with a few discordant positions. I think it was the consensus that was submitted to GenBank (accession CZ473269). One aligns perfectly to 2R: 13839979 (-) (release 6) and that's the best alignment for the slightly mismatched consensus. Of the three 5' flanks, two are identical and align perfectly to the same position as the two 3' flanks (the alignment of this 5' flank sequence against the release 6 genome by FlyBase BLAST is attached as d03281_5_flank_BLAST.png. The 5' flank sequence begins with the 8 bp target site duplication and reads outward into flanking DNA). The third 5' flank is repetitive and aligns with the consensus for TE opus. Perhaps because all three could not be built into a consensus, none of the 5' flanks were submitted to GenBank. Therefore, there is strong evidence that the insertion maps to 2R: 13839979 (-) (release 6), which is 74 bp upstream of the annotated 5' end of the CG6357 gene.