Aligning your experimental run metadata with measurements you’ve taken is a surprisingly time-consuming and fiddly task, particularly for large DOEs with a lot of plates and runs.

Before analysing responses in Synthace or exporting data to a third party tool for analysis you need to use the Prepare Data application to ensure that the factor levels from the design and the measurements you’ve taken are aligned.

The main area is the Data section, which is a table of data allowing you to look at all the data in the dataset (Figure 1).

Figure 1: The DOE Data Preparation App.

Clicking on column headers sorts the data either ascending or descending by that column.

The sidebar to the left allows you to select or deselect columns for inclusion in the data section. Note that these selections only affect the table and are not persisted to the downstream analysis.

The only important control for your downstream analysis is the checkbox marked Arrange replicates into columns. We’ll explain what this does in the next section.

If you have used quasi-replicate factors in your design then it’s likely that your response will be a function of the different measurements taken - for example when comparing a measured signal to a background reading for an equivalent control well, or taking measurements at different time points. (Figure 2)

Figure 2: data before arranging into columns: there is a single A280 column with all replicates appearing as separate rows.

To make this calculation easier in general it’s better to have the measurements for each quasi-replicate in a separate column. Synthace marks replicates at design creation time, so if you have this structure you only need to check the box marked Arrange replicates into columns and the data will automatically be re-structured with replicates in columns. (Figure 3)

Figure 3: data after arranging replicates into columns - note there are now three A280 columns, one per replicate.

To process these replicate columns in Synthace please see the notes in the next section.

In many cases where you use replicates the response you want to analyse is a function of the replicate columns.

Synthace provides some functionality for computing new responses from columns which have been pivoted as part of the Select and Transform tab in the Analyse Responses application.

We provide functions for calculating the mean, variance, standard deviation and also simple linear regression across columns. There is also a generic python expression function which can do much more.

To learn about the predefined transforms see here, to learn about the custom transformation functionality please see here.

Once data preparation is done, press the Start Analysis button in the top-right of the Prepare Data application to proceed to response analysis. (Figure 4)

Figure 4: The Start Analysis button

You will be prompted to give the current version a name, once you have done this the response analysis app will automatically open.

Synthace’s response analysis application provides a simplified interface to data visualisation and DOE model building. To do more sophisticated analyses or more elaborate plots you need to use dedicated data analysis tools.

Specific cases where you will need to do this include

To export the table shown in the data section in xlsx format simply click the Download button at the bottom of the table. (Figure 5)

Figure 5: The DOE structured data download button

Synthace checks your design to see whether it is split across multiple plates, and if so, adds a blocking factor to represent this. This factor is given the name “Plate Blocking ID” and is automatically selected when you structure your DOE data.

Currently, blocking factors to account for plate based effects are the only blocking factors supported by Synthace. There are two situations where the plate-based blocking factor is not added to the data. The first is when Synthace detects that another design factor correlates with the plate. This may be the case when you have split a design across multiple workflows using a design factor. In this scenario, Synthace will display a warning message with a list of the factors that correlate with the plate used:

The second situation where Synthace will not add a blocking factor is when you have arranged replicates into columns. Checking the “Arrange replicates into column” option will cause the plate-based blocking factor to be disabled:

For more information about why blocking factors are important to consider, and how blocking factors impact your analysis, click here.

Well done for making it to the end of this tutorial.

To learn more about response analysis, click here.

To learn more about predefined column based calculations, click here.

To learn more about custom column based calculations, click here.