Independent groups t-test on multiple groups

[z_thiel]

Hello,

since I got excellent help with solving my last problem, I thought I would return with another question:

I performed a screen with a series of variables (each with many replicates), including a negative control. After normalization to the negative control (I figured out how to do that), I would like to calculate p-values using the “independent groups t-test” node between the negative control and each variable and eventually visualize the results in a volcano plot.

I have a table that looks something like this:

variable / value
A / 0.8
A / 0.7
B / 2.1
B / 2.0
B / 2.8
C / 0.3
C / 0.2
C / 0.4
negative control / 1.0
negative control / 1.1
negative control / 0.9

and I would like to transform it into a table like that:

variable / p-value (between variable and negative control)
A / 0.1
B / 0.001
C / 0.001

In the configuration of the “independent groups t-test” node, I can only select two groups so I tried to loop over the individual variables using a grouped loop (as suggested in this thread: Paired t-test). Here, I failed to select the correct “group one” and “group two”, since these always change while the loop is executed.

Do you have any suggestions?

[AlexanderFillbrunn]

Hi @z_thiel,
the trick is to use flow variables in the configuration of the Independent groups t-test node. The Group Loop Start outputs the current group values as flow variables and in the Independent groups t-test’s dialog under the flow variables tab you can assign that flow variable’s value to group 1. I have attached a workflow for you that should get you started.
Kind regards
Alexander

grouped-t-test.knwf (19.5 KB)

[z_thiel]

Yes that is exactly what I needed. Thank you very much!