Key points are not available for this paper at this time.
In addition to the sequence homologies and statistical patterns identified among numerous genetic sequences, there are subtler classes of patterns for which most current computer search methods offer very limited utility. This class includes various presumptive eukaryotic regulatory sites. A critique of the often employed consensus and local homology methods suggests the need for new tools. In particular, such new methods should use the positional and structural data now becoming available on exactly what it is that is recognized in the DNA sequence by sequence-specific binding proteins.
Sadler et al. (1983) studied this question.