The following information was supplied to FlyBase by Marc Freeman, Anna 
Dobritsa and John Carlson in response to a molecular query about the dare gene 
by a FlyBase curator for use in genome annotation.
1. dare[1]
Does the start of accession AF168685 correspond to the P
insertion site in dare[1]?
\**Yes.  We isolated a number of dare cDNAs from a Drosophila head cDNA
library with 5' ends that mapped in genomic DNA 3' of the dare1 P insertion
site.  To determine the true 5' end of the dare transcript (or attempt to),
we performed 5' RACE and obtained a number of products: the 5' end of some
of these mapped to genomic DNA 5' of the dare^1 P insertion site, while
others mapped 3' to the P insertion.  EST LD17269 does indeed map 14 bp 5'
of the P insertion site, and is the longest 5' end from any dare cDNAs, or
RACE products.  At present it is not clear if this is the true 5' end, or
if the length of the 5' end of the dare transcript is heterogeneous, with
LD17269 being the longest version isolated to date.
2. dare[34]
Does the 3' deletion endpoint of dare[34] map to a 0.8 kb
ClaI fragment?
\**Yes. As far as we could tell.
3. dare[5]
Does the 3' deletion endpoint of dare[5] map to a 1.95 kb
PstI fragment?
\**Yes.  This is as precisely as we have defined it to date.
4. dare[4]
What is the 3' deletion endpoint of dare[4]? The P insertion
site of dare[1]?
\**We know that the 3' end of the PlacW element is present in dare^4 based
on Southern analysis.  Everything in the genomic region 3' of the P
insertion
appears to be intact.  We do not know how much of the PlacW element is left
in dare^4, although the 5'-most end of the P insertion appears to be gone.
Our only other information is that the mini-white+ marker has been lost
from PlacW.
We also note that, the 5' end of the dare4 and dare5 deletions have not
been determined.
5. dare[102]
The start of the dare[102] deletion is said to begin 8 bp 3'
of the P insertion site. Do you have a precise endpoint?
\** Yes, we do. The sequence is intact up till the 9th bp after the dare^1
P insertion site. That is the sequence corresponding to the first eight
nts in AF168685 is present in dare^102, then there is a 2054 bp deletion
that removes the entire dare coding region (3' breakpoint corresponds to
the nt 1562 in AF168685). Also, there is a 25 bp insertion of what
appears to be a piece of the inverted repeat of the P element left 
behind (between nt 8 and nt 1562).   
6. rescue construct "GF-AR"
Do the restriction sites of the 6.5 kb BamHI-ClaI rescue
fragment correspond to the first BamHI site and second ClaI site on
the map in figure 2A? (I realize that a 0.5 kb SacI fragment was
removed too.)
\**Referring to figure 2A: GF-AR begins at the second (not the first) BamHI
site, and extends to the second (as you stated above) ClaI site.
7. dare transcript
The first intron does not seem to be spliced out of EST
LD17269: do you think it is simply an unprocessed transcript?
\**We have sequenced across the first intron in genomic DNA, as well as two 
independently isolated cDNAs.  Both show the same splicing pattern as diagrammed 
in Fig. 2A and indicated by the cDNA sequence (AF168685) when it is compared 
to genomic DNA.  The protein product predicted from this splicing pattern aligns 
very well with bovine, human, and yeast AR, and is therefore likely to be correct.  
(We also rescued many phenotypes with a hs-AR construct that had a fragment
corresponding to AF168685.)