Track standard dilutions responses from pools of 6, 12, 20 and 30 assays.

• Are there non-monotonic regions, or hooks, at either end?

• What is the variation between the dilution responses of multiple assays?

• What is the dose range of dilution responses that has the most reliability?

Standard Dilution Response Limits Are Established For You

• The graph on the left shows the responses from each of the standard dilution series in your pooled assays.

• The graph in the middle shows the confidence limits for each standard dilution using the data from your pooled assay responses, after filtering out any outliers.

• The graph on the right shows these limits plotted with the standard dilution responses (red dots) of your assay. From this graph, you can see how the assay is behaving compared to the expected behavior determined by your previous assays.

Distribution Of Standard Dilution Responses

• Is the distribution of responses from each dilution a normal (bell shaped curve) distribution?

• Do any end dilutions show no little or no spread in its distribution, indicating reaction or signal saturation at those doses?

• Are there any outlying dilution responses from the pooled responses that warrant further investigation of an assay?

Intra-Assay and Inter-Assay %CVs

• The intra-assay %CV between replicates (open squares) and the inter-assay %CV between assays (filled squares) from the pooled assays of the baseline, standard, and control dilutions should be uniform for all dilutions.
• Higher %CVs at either end indicate the limits of the useful dose response curve. Dilutions with higher response %CVs will be more noise with little contribution from the assay’s biological kinetics, as demonstrated by the blank baseline (Bn) and first standard dilution responses shown in the graph above.
• The control dilutions should have intra-assay and inter-assay %CVs similar to the standard dilutions at those doses. Differences in %CVs indicate matrix differences between the standard and control dilutions.

Standard Dilution Response Residuals²

• The individual residuals² (red bars) are the pooled assay averages of the squared differences between the observed and predicted response off the curve of each dilution, divided by its expected variance (weight).

• The sum of these residual² averages should equal the degrees of freedom (number of points – number of curve parameters).

• The distances between adjacent standard dilutions should be relatively uniform.

• There should be no clumping of responses at either end. Clumped responses have higher than average residuals² and worsen the fit and accuracy of the regression curve. Clumped responses at the ends often indicate nonmonotonic hooks.

Distribution Of Weighted Standard Dilution Residuals²

• The individual residuals² (red bars) are the squared differences between the observed and predicted response off the curve of each dilution, divided by its expected variance (weight).

• The distribution of these weighted residuals² should be random without a single pattern for one dilution across all of the assays and the opposite behavior for an adjacent dilution.

• The weighted residuals² of individual assay dilutions will reflect the distribution of curve RSSEs of the assays.

Standard Curve Precision Error Profiles

• STATLIA MATRIX generates precision error profiles (red lines) of the concentration errors of the pooled standard curves, plotted against the desired reportable range (green bar). See the manuscript on the software’s error profile method.

• These plots shows the limits of quantitation at any desired concentration error threshold under the protocol conditions of the pooled assays.

• Precision error limits of quantitation often can be expanded by changing standard dilution doses or other protocol conditions.

• The precision error profiles allow you to gauge the usability of the test protocol against the desired reportable dose range of a quantitation test or effective dose range of a potency test.

Monitor How Your Data Are Trending Across Assays

• STATLIA MATRIX plots the responses of each baseline, standard, and control dilution for each of your pooled assays so you can track the trending of your responses from assay to assay, as shown in the first graph. Individual graphs are plotted for your raw responses, your responses adjusted to remove label variation when a tracer is assayed, and your responses normalized to your highest response dilution, and your Y-axis responses.

• The results from each individual control specimen are also plotted for each assay, as shown in the second graph, along with each control’s respective confidence limits computed from their pooled assays or established by the laboratory.

• The individual dilutions and control specimens are all color coded so they can be easily distinguished.

Detect Assay Drift Over Time

• Most test methods are stable over time. But sometimes reagent lot changes or other factors can cause assay behavior to change. STATLIA MATRIX automatically monitors the behavior of a test over time and creates detailed Quality Assurance reports of this behavior periodically, so you are alerted to any behavioral changes.
• In the above graphs, the green regions are from the reference assay pools, the red regions are from the new assay pools, and the rust regions are the overlap of the two pools. The new assay pool from the test method on the left shows little drift from its original reference assay pool. The same test method on the right shows an extreme case of behavior change in assays that used a new lot of a critical reagent.
• STATLIA MATRIX tracks the reagent lots used in each assay in its Reagent Inventory. You can use STATLIA MATRIX’s Quality Assurance feature to compare assays that used a new reagent lot to show regulatory compliance.