Nevertheless, it is being presented in this paper as it is applicable to analyzing any similar sigmoidal curve relationship in Excel, which is almost universally used. Furthermore, although the template provided here will work satisfactorily in the majority of cases, savvy users may modify the formulas and VBA code to suit their particular circumstances more precisely. However, the results provided by the Excel template are restricted to the regression line and the estimates of c and d of Eq.  (1), and do not permit the response of the flies to the anesthetics to be classified into sensitive, normal or resistant types — one of the major goals of the laboratory. The stand-alone

GUI-based this website Windows program HEPB does the same analyses as above, but in addition it constructs a prediction band at a user-defined confidence level and then determines the cut-off values from those prediction band limits that help to objectively distinguish among sensitive, normal and resistant phenotypes. These values also enable GSI-IX in vitro researchers to determine rapidly and objectively if experimental values are statistically different from their control ranges in their assays. As far as we are aware, HEPB is the only program that does

the four-parameter logistic regression, constructs the prediction band for the data, and provides objective, empirically determined cut-off values to distinguish among response phenotypes. Furthermore, it optionally generates 500 simulated values of the response variable within the range of the observed dose variable. This can be useful particularly when the sample size is limited and the user is unable to visualize the dose–response behavior in the data. While it might seem redundant to provide these two different avenues for performing this analysis,

much we believe that each program fills a niche within the laboratory. Most users will find the Excel template straightforward and will be comfortable with its interface. Additionally, it can interface with other Microsoft Office software, like Access, to store data in a laboratory database, if needed. There are other sources that also involve the use of Solver to fit non-linear equations (Harris, 1998). In addition, there are instructions available in several websites on the internet. However, none of these sources provide a template such as the one presented here that not only makes it easy for the uninformed user (who merely needs to enter the data in the template) but more importantly, has been programmed to auto-check for the goodness of fit and redo the analysis with sets of alternative starting values for c and d in Eq.  (1) until the goodness of fit criterion is met. It has been tested with a number of datasets that span a wide range of relationships between the dose and response and sample size ( Fig. 9), and has performed remarkably well ( Table 1).