Algorithm Constants for Determining the Shape of the Standard Curve
In addition to the straightforward parameters such as the assay name and the desired units to be output, several key constants must also be resolved prior to creating a new assay. These constants determine the shape of the standard curve used in calibration and are identified in Table 1. Depending on your interest level and application type, this may require collection of new data.
In this case, the Qubit® 2.0 Fluorometer can be used as a mini-fluorometer in Qubit® Raw mode (right click, save target as, save to desktop) This function allows you to manually select the excitation light source of your choice (i.e., blue LED, red LED, both, or none), while reading fluorescence in both the green and far red emission channels. Raw fluorescence values are obtained to generate data for the development of new assays, or to validate the performance of existing ones. Raw fluorescence values are reported to one decimal place.
Different constant values can be compared with similar data sets to determine the values that result in the best accuracy. Since the curve-fitting algorithm accounts for some sigmoidicity at the upper and lower portions of the standard curve, we have routinely found that results obtained on the Qubit® 2.0 Fluorometer are more accurate over a broader dynamic range than results obtained using a traditional fluorescence microplate reader and a simple linear fit using Microsoft® Excel or similar data analysis software.
Example of Optimizing Algorithm Constants
As an example of how to collect data and optimize algorithm constants for new assays created using MyQubit, we have provided the following Microsoft® Excel file outlining an experiment performed with the Amplex® Red Cholesterol Assay Kit . At the top of the page is the general algorithm, based on a modified Hill plot, used in the creation of all new assays. The constants required in the .qbt file, as described in Table 1, are highlighted in blue. In this case, according to the procedure outlined in the amended Qubit® cholesterol assay protocol , triplicate samples of serially diluted cholesterol reference standards from concentrations of 0.3125 to 20 μM were incubated for 15 minutes at room temperature and read using Qubit® Raw mode (after addition of the Amplex® Red/UltraRed Stop Reagent. Concentration values were determined for each sample based on its raw fluorescence value and the displayed algorithm, and were compared to the known concentrations. Near the bottom-right portion of the file, you can see a summary of average deviation and accuracy results obtained across the entire sample range using the highlighted algorithm constants. You may wish to change one or more of the constant values in this file to see how the calculated outputs are subsequently affected.