More Arabidopsis sequencing

Barbara Baker bake at ENZYME.BERKELEY.EDU
Thu Jan 30 00:56:05 EST 1992

> Rob Last noted that it would be useful to place sequenced cDNAs on the map.
> We certainly agree that in many cases this will be the only way of gaining
> insight into what a cDNA encodes. We think the most efficient way of doing
> this will be to hybridize YACs (from the forthcoming set of overlapping YACS)
> to the grid of cDNAs.  For this to work, there must no be any homology between
> the cDNA cloning vector and the YAC vector.  We currently interpret this to
> mean that the cDNA libraries should be in lambda vectors that can be sequenced
> by PCR methods.  I would be interested in hearing better ideas. Chris

Yes. I believe the reverse approach would be better, hybridizing YAC grids
with labelled cDNAs. 1) It is easier to isolate probes from the cDNAs than
YACs (for example, by PCR amplification of bacterial colony lysates or
plaques). I think the short,oligonucleotide sequences at the ends of these
PCR products would not equal signal generated by hybridization of the cDNAs
themselves, should the polylinker sequences be present on the YACs as well.
2) Perhaps the YAC libraries will be complete and gridded before a large
number of cDNAs are available as clones, and sequenced (the latter set
grows incrementally, the former more immediately?). 3) What we don't
know is whether these YACs are ordered or not. If they are not, then
hybridizing them to cDNAs serves a dual function : mapping the cDNAs 
(assuming they are not familial in stringent conditions) and helping order
the YACs into contigs. If the YACs are ordered then only mapping cDNAs is
accomplished. It would seem that the utility of using cDNAs to establish
contigs is a function of the size of the library - more overlap, more
information. Little overlap, little chance that a sequence finds
more than one YAC. Interesting...                                Brian Osborne

More information about the Arab-gen mailing list