cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Check out the JMP® Marketplace featured Capability Explorer add-in
Choose Language Hide Translation Bar
MSlobodian
Level II

Having trouble putting multiple groups into the same LC50 graph.

Hello,

 

I am a new JMP user so I am very unfamiliar with the program. I just conducted an LC50 experiment where I compared the lethality of two different toxins at a range of concentrations. We also separated our flies by sex and genetic background. We're able to create the curves, but we can't figure out how to have multiple groups on the same graph. For example, we want to create a curve which has the log of our concentration on the x and % Dead on the y, with 8 curves on the graph each representing one of our four genetic background and one of the sexes (2 sexes by 4 genetic backgrounds). When we try to use the fit curve function, we couldn't input more than one group, and when we tried using Fit Model instead, JMP wouldn't produce a graph if we used more than 2 columns in the constructed model effects. We can easily create a graph with either our 4 genetic backgrounds or our 2 sexes, but never with both on the same graph.

 

Thank you for the help.

1 ACCEPTED SOLUTION

Accepted Solutions

Re: Having trouble putting multiple groups into the same LC50 graph.

Would a probit analysis suffice? If so, then set up your data table so that you have a separate data column to indicate group membership for each grouping, a data column with the total number of subjects, and another data column with the number of dead subjects. Here is an example from a JMP training course where there is only one grouping variable, Pressure.

 

markbailey_0-1635946772804.png

 

 

Set up the analysis like this:

 

probit.PNG

 

Note that the order of the data columns are entered into the Y role.

 

You now get this result.

 

markbailey_1-1635946801804.png

 

The Profiler can be used to find LC50 and other information

 

markbailey_2-1635946846190.png

 

.

View solution in original post

4 REPLIES 4
Thierry_S
Super User

Re: Having trouble putting multiple groups into the same LC50 graph.

Hi,

It is rather challenging to provide clear guidance without seeing the current format of your data table, but I would suggest that you stack your data with a single column for X, % Dead (i.e., Y), and a grouping variable form by the concatenation of the Sex + Genetic background. Then, you can plot the %Dead by X with the Grouping variable as Overlay in GraphBuilder.

Does that make sense?

Best

TS

Thierry R. Sornasse
MSlobodian
Level II

Re: Having trouble putting multiple groups into the same LC50 graph.

Just tried that out and it worked! Thank you so much for the suggestion!

Re: Having trouble putting multiple groups into the same LC50 graph.

Would a probit analysis suffice? If so, then set up your data table so that you have a separate data column to indicate group membership for each grouping, a data column with the total number of subjects, and another data column with the number of dead subjects. Here is an example from a JMP training course where there is only one grouping variable, Pressure.

 

markbailey_0-1635946772804.png

 

 

Set up the analysis like this:

 

probit.PNG

 

Note that the order of the data columns are entered into the Y role.

 

You now get this result.

 

markbailey_1-1635946801804.png

 

The Profiler can be used to find LC50 and other information

 

markbailey_2-1635946846190.png

 

.

MSlobodian
Level II

Re: Having trouble putting multiple groups into the same LC50 graph.

The Prediction Profiler is really useful! I didn't know that was a feature. Thank you so much!