The rapid pace at which the human genome project has proceeded has gre
atly benefited from two classes of short sequence tags, genomic (STS)
and transcribed (EST), which are listed in two separate databases. Usu
ally, STSs are random genomic sequences derived only for mapping purpo
ses, while ESTs represent transcribed sequences that have to be mapped
one by one. Here, we propose a way of establishing links between thes
e two sets of sequences, allowing the automatic mapping of EST sequenc
es by simple comparison with relatively nonrandom STSs. We suggest tha
t EagI-based STSs derived by selected genomic portions organized in YA
C contigs can automatically finely map a relevant portion of the ESTs,
partially bridging the gap between the two sets of sequences and savi
ng a great amount of time in mapping efforts. To test this principle,
we have selected 330 high-quality STSs derived from the Xq24-qter regi
on and used them for transcript searches by comparing them to the EST
as well as to the nonredundant database. This search detected four kno
wn genes and two additional EST clones. In contrast, when the same dat
abases were searched with a set of 53 sequences derived from the same
chromosomal region around EagI sites, 7 known genes and 6 additional E
STs were found. These findings, together with data obtained from simul
ation analysis on long sequences in the same chromosomal region, sugge
st that EagI-based STSs can partially bridge the gap between STSs and
ESTs. (C) 1996 Academic Press, Inc.