In addition to the following excerpt from the NONLIN manual, see ref 4: Johnson, Michael L. and Lindsay M. Faunt (1992) Parameter estimation by least-squares methods. Meth. Enzymol. 210:1-37. In this article a comparison is made of the most rigorous to least rigorous methods of confidence interval estimation. The method used by NONLIN is a compromise between rigor and computer time needed to conduct the analysis, but is still better than many approaches that completely fail to deal with covariance between parameters.

Excerpt from NONLIN Manual:

The reported confidence limits are calculated by searching the variance space for an F-statistic corresponding to approximately a 65% confidence probability (Ackers, et al., 1976; Johnson et al., 1976; 1981). The 65% confidence region for a Gaussian distribution is the mean plus or minus roughly one standard deviation. This search of the variance space is performed in two ways: In the first way, each of the elements of A is varied independently. In the other way, the direction of the search is defined in terms of the F statistic (variance ratio) as the axis of the multidimensional hyperellipsoid defined by the solutions, O, of the following matrix equation (Box, 1960; Magar, 1972; Johnson et al., 1981). The use of option 1 is recommended.

       (A-O) P'P(A-O) < n v (F statistic),

where n is the number of parameters and v is the variance. Confidence intervals evaluated by this procedure correspond to approximately 1 standard deviation, but because of the correlation between successive data points and between parameters, these confidence limits are only estimates of the true value. In general such confidence limits will be asymmetrical and are thus reported as a range of values instead of a single value. These confidence limits only reflect the precision of the fit of the experimental data to the model and do not necessarily indicate the accuracy of the determined parameters. The evaluation of the confidence region does not include possible effects of systematic errors in the data. The use of option 1 is recommended. For further detail, see [Eq.49] (Straume, Frasier, and Johnson).

Straume, M., Frasier, S.G. and Johnson, M.L. (1988), Least-Squares Analysis of Fluorescence Data, in Fluorescence Spectroscopy: Principles and Applications (J. Lakowicz, ed.), Plenum Press, New York, in press.

Return to calling text characterize the physical basis for signal transduction from the N-terminal domain to the rest of the activator protein. We also collaborate with Dr. Tim Hoover (Microbiology, University of Georgia) who is examining the physical basis for interaction between the activator and the RNA polymerase.

We also collaborate with Dr. Dale Kaiser of the Department of Microbiology at Stanford to examine the role of similar activator proteins in controlling the starvation response of Myxococcus xanthus, and with Dr. Karen Miller of the Department of Food Science at Penn State to investigate the role of cyclic-a-glucans in the nodulation process that leads to effective symbiosis between rhizobium species and legumes.

Representative Publications

Nixon, B. T., C. W. Ronson, and F. M. Ausubel. 1986. Two-component regulatory systems responsive to environmental stimuli share strongly conserved domains with the nitrogen assimilation regulatory genes ntrB and ntrC. Proc. Natl. Acad. Sci. USA. 83:7850-7854.
Miller, K. J., M. W. McKinstry, W. P. Hunt, and B. T. Nixon. 1992. Identification of the diacylglycerol kinase structural gene of Rhizobium meliloti 1021. Mol. Plant Microbe Interactions. 5:363-371.
Ledebur, H., and B. T. Nixon. 1992. Tandem DctD binding sites of the Rhizobium meliloti dctA UAS are essential for optimal funciton despite a 50 to 100-fold difference in affinity for DctD. Mol. Microbiol. 6:3479-3492.
Gu, B., J. H. Lee, T. R. Hoover, D. Scholl, and B. T. Nixon. 1994. Rhizobium meliloti DctD, a e54-dependent transcriptional activator, may be negatively controlled by a subdomain in the C-terminal end of its two-component receiver module. Mol. Microbiol. 13:51-61.
Lee, J. H., D. Scholl, B. T. Nixon, and T. R. Hoover. 1994. Constitutive ATP hydrolysis and transcription activation by a stable, truncated form of Rhizobium meliloti DctD, a e54-dependent transcriptional activator. J. Biol. Chem. 269:20401-20409.