I am writing to alert researchers to a possible artifact in the mapping of three Exelixis piggyBac insertions within the Rbp1 gene.  The Exelixis collection of insertion mutants, described by Thibault et al. 2004 (FBrf0174227) were mapped by Exelixis Corp by sequencing inverse PCR products spanning the transposon-flanking DNA junctions.  Members of the Gene Disruption Project subsequently reanalyzed the Exelixis sequence data and deposited flanking sequences in GenBank (FBrf0184340 and manuscript in preparation). Three insertions of the piggyBac RB vector were mapped within an intron of the Rbp1 gene: e01298, e04474, and e02960.  For e01298 and e04474, only the 3prime flank of the transposon was recovered; for e02960, the 3prime flank mapped to Rbp1 on chromosome 3 and the 5prime flank mapped to a site on chromosome 2.  I am concerned that the mapping of these insertions to Rbp1 may be an artifact because the RB transposon itself contains this segment of Rbp1 and !
the putative flanking sequence aligns entirely with this sequence in the vector.  The sequences of e01298 and e02960 have a canonical piggyBac 3prime end sequence joined to a sequence of Rbp1. This could be interpreted as either a real insertion of the piggyBac into Rbp1 or a transposase-induced recombination joining the 3prime end of the transposon to an internal sequence within the transposon itself.  The sequence of e04474 provided by Exelixis is low quality in the region of the vector-flank junction and I can't see whether there is a canonical piggyBac 3prime end sequence joined to the higher quality Rbp1 sequence.