To determine the frequency and clustering of a variety of simple di- a
nd trinucleotide repeats, an Artiodactyl short interspersed element (S
INE), an ovine satellite repeat, and a human Alu 1 repeat were used to
screen a random selection of cosmids containing inserts of ovine geno
mic DNA. In total, 197 individual cosmids were digested with EcoRI and
the fragments separated on 0.7% agarose gels. Southern blots of these
gels were then sequentially probed with (AC)7, (CT)9, and (CAC)6 olig
onucleotides, and the repeats described above. The frequency at which
(AC)n, (CT)n, and (CAC)n repeats were found in the cosmids indicated t
hat they occurred at average intervals of 65 kb, 367 kb, and 213 kb re
spectively within the ovine genome. The Artiodactyl SINE was the most
common, occurring at an average interval of 20 kb. No human Alu 1 sequ
ences were detected. There was a significant positive association betw
een the (AC)n and the Artiodactyl SINE. This association is quite stro
ng as there was significant clustering of the two repeats both within
cosmids and also within the EcoRI fragments of the digested genomic fr
agments. With the exception of the sheep satellite sequence, which occ
urs in tandem arrays, none of the other repeats showed significant clu
stering within the 41-kb (average size) cosmid inserts. The first 25 o
vine microsatellites we characterized had an average polymorphic infor
mation content (PIC) of 0.65. The different microsatellite types, cont
aining either perfect, imperfect, or compound repeats, had similar ave
rage PICs of 0.64, 0.65, and 0.66 respectively. There was a weak regre
ssion relationship (R2(adj)% = 21.9) between the length of the longest
uninterrupted dinucleotide repeat in the largest allele and the PIC o
f the microsatellite.