On 3/11/14 3:41 PM, pfotiad@nmr.mgh.harvard.edu wrote:
Hi Doug,
Thanks for your reply. Just to be clear, I need to ask some trivial questions:
- I need to do that subtraction in the fsgd file right? (In which case
there will be negative numbers as well)
Correct.
- I ran the analysis and the resulting significant clusters when looking
at the results of [1 -1 0 0] look very similar to the ones that I got when I ran the analysis without accounting for age (that is an fsgd file with no age included and a design matrix of [1 -1]), but not that similar to the results that I got when looking at age=0. That makes sense, right?
I don't know. If there is an interaction between age and group it becomes hard to predict what should happen if you do not include age as a covariate doug
Thank you for your time, Panos
On 03/07/2014 07:46 PM, pfotiad@nmr.mgh.harvard.edu wrote:
Hi Doug,
Thanks for your reply. Yes there were areas of significant difference reported in the age slopes maps, after the mri_glmfit command was run. However, I haven't checked to see yet whether there are surviving clusters after correction for multiple comparisons. I see, yes that definitely makes sense especially since the slopes have an opposite sign. However I still had a couple questions:
- If I did want to look at what the age-corrected trait looked like
when compared between the Diseased and Healthy groups at age=mean instead of age=0, what would I need to do?
Subtract the mean age (computed across all subjects regardless of group) from the age of each subject
- When not including the age in the fsgd file and using a design matrix
of [1 -1] I get the significant differences between the average trait in the diseased cases vs the average trait in the control cases. Is there a way to get a summary file with the actual averages along with the significance maps?
If you run mri_glmfit-sim it will produce a lot of summaries. Run it with --help to see if what you want is there doug
Thanks a lot for your help! Panos
Was the difference in the age slopes significant? Ie, and interaction between age and group. If so, then you have to proceed with caution (at least one statistician told me not to proceed at at:) because the difference in the intercepts between the two groups will depend on age making it very difficult to interpret. At the age where the lines intercept, there will be no difference doug
On 03/07/2014 02:57 PM, pfotiad@nmr.mgh.harvard.edu wrote:
Hi Doug,
I ran a group analysis comparing a trait between two groups (Diseased vs Controls) with one continuous variable (age), with the following design matrices [1 -1 0 0] and [0 0 1 -1]. When I loaded the significance map and the descriptor file, the age slope of the trait in Controls was positive whereas the age slope of the trait in the Diseased cases was negative. Therefore, comparing the intercept between the two groups does not really give that much info, since the slopes are of opposite sign. What do you believe would be the best course of action in order to compare the age-corrected traits:
- instead of age=0 maybe comparing the groups at age=mean?
- or going with a DOSS model?
Thank you for your time, Panos _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer