Tips for Data Analysis in the Q-View™ Software

The Q-View™ Software is a tool for the quantitative analysis of multiplexed chemiluminescent or infrared fluorescent assays, such as those found in Q-Plex™ planar-based arrays. The software enables users to:

• Acquire images of microarrays and stack images optimized for bright and dim reactions, obtaining a single high dynamic range image.
• Easily locate spots for each assay using the software’s Auto-Set Plate Overlay and Auto-Adjust Spots features.
• Assign wells as samples, controls, standards, or negatives, and specify their dilution factors.
• Fit curves using any of six curve-fitting models, including 4 and 5 Parameter Logistic (PL), and customize or export the automatically generated charts and reports.

How does Q-View calculate concentrations?

Once the plate overlay has been placed, foreground pixels within each overlay circle are averaged using a proprietary weighting algorithm which is designed to prevent well background from having a large impact on the calculation.

How do I know what curve fitting model to choose?

We recommend the 4PL or 5PL regression models for Q-Plex data, as these are generally considered to provide the best overall fit for immunoassay curves. Use the Auto-Select option to have the Q-View Software automatically fit each curve using the model with the lowest AIC statistic. If your data is abnormal or you are trying to optimize the fit in a particular range (see below), one of the other Q-View regression models may better fit your needs.

When should I mask a standard curve point?

Consider masking a standard curve point if the %CV between the standard curve replicates is greater than 20%, or if the average %Backfit is outside 80% – 120%. You might also consider masking if you are trying to optimize the fit in a particular range (see below).

Can I optimize the curve fit for the low or high end of the curve?

Yes, in Q-View, try masking wells at the other end of the curve and choosing either the Auto-Select or Log-Log option. For example, masking the top 2-3 points of the standard curve and using Log-Log may better fit the low end of the curve.

Can I report pixel intensities?

Maybe. There are some issues with reporting pixel intensities, including:

• Data is no longer “normalized” between plates such that small differences in things like the SHRP incubation time or time from when substrate is added until the plate is imaged can cause significant differences between experiments.
• The fact that the signal is non-linear (due to both kinetic and camera effects) is no longer taken into account.
• Some reviewers will not accept data below the assay Lower Limit of Quantification (LLOQ) because it is considered likely to have unacceptably high variability, recovery, linearity, etc.

Some reviewers will accept Pixel Intensities if the data is presented as qualitative rather than quantitative. For example, if you could establish that the Pixel Intensities of your samples were statistically different than your negative controls, then you may be allowed to report Pixel Intensities relatively. One convention for doing this is to use “the sum of 6-10X the standard deviation of the negative controls plus the average negative control value” as the cutoff point.