MBA output

AFNI Message Board
1

Hi Gang

I am running MBA on a control vs patients study and I have a table setup this way:

Subj ROI1 ROI2 Y tp
103 Left_Motor Right_Motor 0.9057 control
…
sub1 Left_Motor Right_Motor 0.7641 pat

Now if I run the model with this call
MBA -prefix cont_post -chains 14 -MD -iterations 5000 -model ‘1+tp’ -cVars ‘tp’ -EOI ‘Intercept,tp’ -r2z -dataTable …/controls.txt

I get the Intercept.pdf that is fine, but also another file called ‘tp_controlsvspre.pdf’ where the ROIs have been replaced by numbers.
I wanted to ask what this output represents and if the ROIs have the same order as in the Intercept pdf.

thanks

2

another file called 'tp_controlsvspre.pdf' where the ROIs have been replaced by numbers.

Strange indeed. In Intercept.pdf you do see those region labels?

You don't really have to run the analysis with 14 chains. Usually 4 (at most 8) would be enough. Do you really need 5000 iterations?

3

Thank you for your reply.
Yes I see the labels in the intercept. I am attaching both files to show this finding. I am using an updated version of afni (downloaded last week).
Does this file represents just the difference in effect magnitude of the levels specified by the variable ‘tp’ in my case?
Thanks again

tp_controlsvspre.png
tp_controlsvspre.png800×533 85.4 KB

Intercept.png
Intercept.png804×529 97.1 KB

4

Does this file represents just the difference in effect magnitude of the levels specified by the variable 'tp' in my case?

Yes, that's correct. The labels should be the same (and in the same order) as the intercept. I may have to ask you to send me the data to diagnose the problem.

5

Thank you a lot Gang for your help. I have attached the datatable at this link as a text file https://osu.box.com/s/4o6qb96icff6sgpc26y4mc3i7s90dbnf
Let me know your thoughts
Thanks again

Best regards

6

Thanks for the data. The fix for the labels will be in the next AFNI release. Let me know if you want the fix earlier.

7

Thank you would the fix be already available in release AFNI_19.1.11 ?

8

I’m not so sure about the AFNI_19.1.11 version, but the next one AFNI_19.1.12 in the distribution should be fine.