One promising technique for determining long genetic sequences is sequ
encing by oligonucleotide content. This technique involves probing a s
egment of the unknown multimillion ''character'' genetic sequence for
the presence or absence of known short subsequences. The information o
btained from such hybridization experiments may be represented in netw
ork form. Network optimization methods may then be applied to identify
the most likely forms of the unknown target sequence. (C) 1994 John W
iley & Sons, Inc.