INTRODUCTION
Experimental data on site activity are well represented in the literature. McClure and co-workers were the first to accumulate data on natural E. coli promoter strength and to apply them for the prediction of the strength of some other natural promoters (Mulligan et al., 1984). Later, Jonsson et al. (1993) on the basis of experimental data on natural E. coli promoter strengths have developed a method applicable to the analysis of artificial point mutations in natural E. coli promoters. Berg and von Hippel (1987, 1988) have collected data on the activities of prokaryotic activator and repressor binding sites, these data being the foundation for the commonly accepted statistical-mechanical theory of DNA-protein interactions. Stormo and co-workers (Stormo et al., 1986; Barrick et al., 1994) were the first to apply a wide spectrum of mutational events, mutagen-dependent hot spots, nonsense codon suppression, ribosome binding sites to sequence-activity analysis. Kraus et al. (1996) have initiated investigations in eukaryotes: they studied transcription initiator (Inr-element) and TATA-box activities and predicted them successfully. All these studies were made assuming that sequence-activity relationships are invariant, thus, ignoring the conditions of the measurement system.
At the same time, Sarai and co-workers (1998) have observed that point mutations in the OR1-operator act differently while binding with two antagonist proteins, Cro- and l -repressors. Analogously, point mutations in the c-Myb binding site cause different free energy changes (D D G's) in natural (Tanikawa et al., 1993) and mutant target proteins (Ogata et al., 1996). Analogous data were obtained for the natural, EREBP-2, and homologous proteins, EREBP-4 and AtEBP-1, binding the GCC box in plants (Hao et al., 1998). Similar evidence was obtained by comparing in vivo and in vitro systems of the Inr-element (Kraus et al., 199) and TATA box (McCormick et al., 1991) activities. Analogously, Hyde-DeRuyscher at al. (1995) have found a discrepancy in activity measured in different cell systems and plasmid constructs. Also, Javahery et al. (1994) have found no correlation between the Inr/YY1-induction magnitudes and YY1/Inr-affinities.
In this respect, currently the ACTIVITY system accumulating experimental data on DNA/RNA site activities and sequence-activity relationship at fixed experimental conditions links to the SYSTEM database, accumulating measurement systems, and the CROSS_TEST database, accumulating procedures for the application of one measurement system to another.
ACTIVITY are useful for molecular biology, pharmacogenetics, metabolic engineering, drug-design, and biotechnology.