Hi, my name is Florian Müller-Sallanz and I'm a statistican at Stegmann Systems. In today's article, I'm gonna show a neat trick regarding the use of our Control chart documents for Statistical Process Control.
The great workshop by Nancy Niemuth and Laureen Little at last year's BEBPA in Budapest emphasized yet again the recent surge of interest in employing SPC techniques such as WECO for monitoring assays. The formulas to compute the relevant statistical values and decision rules are easily implemented. More often than not, they are already included in modern spreadsheet programs.
But have you ever tried to compute a mean and an estimated standard deviation by manually transferring the value of the EC50% or the value of the slope for, let's say, the last 100 assays? Tedious and prone to errors...
Why not use the Control chart documents of the Bioassay Package?
Hidden Features of the Control Chart
Control chart documents offer various methods to visualize the evolution of your assay. They can also utilize control limits to implement process control. And although they do not yet support full-scale SPC, you can already today export parts of your assay data to your favourite CSV editor or spreadsheet program such as MS Excel, perform SPC, and analyze and visualize your historical assay data.
As you know, the C Parameter of the Standard is closely linked to the Standard's EC50%, so in the following example, i'll show you how to check the sigma-limits of this Parameter.
Setting up the Document
Start by including the relevant values in your chart. In the screenshot below, the Control Chart automatically includes the C parameter of the reference sample of all the assays in the source folder.
To learn more about how to configure the dataset of a Control chart document, check out LINK. So far, everything is as usual: after updating the references, you are now ready to create the chart, but you won't have any limits yet.
Augment the Chart with the Power of Spreadsheets
To perform additional analysis through a spreadsheet, switch to the 'References' editor of the Control Chart document.
Now click 'Open Externally'. This opens the dataset in your default CSV editor.
Now you can plot or analyze the data to your heart's content. In our example, we're interested in the mean, the standard deviation and the sigma-bands.
Please note that values are initially exported with the precision displayed in the table of references, but depending on your application some decimals might be discarded upon saving or might not be displayed.
Back to the Chart
To conclude this example, we transfer the sigma-bands back to the control chart to supplement the visualization.
Now we can quickly verify if any WECO-Rules where broken by our assays. Rule 4 was violated once by assay runs 3 to 12, but everything else is looking fine.
Tl;DR : While the Control Chart document does not yet support complex statistical analysis of the charted data, exporting the values of the dataset to a .csv via 'Open Externaly' enables you to analyze, trend or plot various values of you historical assays.