Subject: Alleles changes based on Levis pc
Personal communication to FlyBase from Robert Levis
March 2003
 Note:In  cases in which the flanks map to different nearby sites, there is
no way to determine from invPCR data whether there are two insertions
or an insertion associated with a deletion. I would think that it
would be simplest to record it as two insertions, but put in a caveat
in a comment field to the effect that this could be two insertions or
one insertion associated with a deletion
KG00307
Previously reported to be multiple insertions 14 nt apart. This is
incorrect. Flanking sequences map to the same site. The 3' flank begins
with a 12 nt sequence that does not align with genomic at the insertion
site. This sequence is found in the vector. I'm not sure whether this
is an invPCR artifact or a sloppy P element insertion (that added extra
nucleotides from the P-element), but it does NOT appear to be a double
insert.
KG00799
2 insertions within Dscam; flanking sequences map 169 nt apart on opposite
strands.
KG06827
2 insertions within Hnf4; flanking sequences map 11 nt apart.
KG00542
2 insertions; flanking sequences map 2415 nt apart on opposite strands.
The 3' flank is in the 5' UTR of PGRP-LF (CG4437) and 5' flank is in
CDS of CG4347.
KG01370
Previously reported to be multiple insertions 18 nt apart. This is
incorrect. Flanking sequences map to the same site.
KG03309
Two insertions 2.5 kb apart or one mael insertion and a 2.5 kb deletion.
KG03939
2 insertions; flanking sequences map 484 nt apart on opposite strands.
KG01645
No evidence for more than one insertion.
KG05879
2 insertions; flanking sequences map 72 nt apart.
KG01037
No evidence for more than one insertion.
KG01248
There is marginal evidence for a 2nd insert at an undetermined site.
The 5' flank is 139 nt long and aligns at the site recorded for
KG01248a. The 3' flank was 29 nt long in the original and 33 nt long
in the recheck (the 29 nt is a subset of the 33 nt). The 3' flank
and 5' flanks do not have 8 nt target site duplication. The 3' flank
is not only short, but has a low complexity because of a run of 13
T's. When I BLAST it vs all Dros sequences, the 3' flank does not
give any really good alignments.
KG00877
2 insertions; flanking sequences map 16 kb apart, 5' CG17741, 3'
unknown.
KG03323
No evidence for more than one insertion.
KG00779
No evidence for more than one insertion.
KG00872
Previously reported to be multiple insertions 24 nt apart. This is
incorrect. Flanking sequences map to the same site.
KG02394
No evidence for more than one insertion.
KG02862
Short 5' flank that mapped to 2L, but didn't reappear on recheck.
Perhaps this should be entered as a single insert with a comment
that there is a possibility of a 2nd insert at 2L-21E4-21F1.
KG03784
No evidence for more than one insertion.
KG04515
No evidence for more than one insertion.
KG05115
No evidence for more than one insertion.
P{GT1}cherBG02734a
The flanks of BG02734 map 8 nts apart on opposite strands.  Perhaps a
tandem double insert?
P{SUPor-P}KG01280a
No evidence for a 2nd insertion for KG01280.