Dear Doug,
thank you very much for your advice. I re-ran the contrast to look at age-area differences between the two slopes. There were no significant differences after monte carlo simulation so I changed the mriglmfit script to doss, as you suggested (but left the .fsgd file as is).However when I do this and re-run the simulation to look at genotype differences, my significant results that I previously found disappear. Why should this happen as there is no difference between age-area slopes? Do I need to changemy .fsgd file also?
Thank you again for your help,
Sinead
On 5 August 2013 05:05, Douglas Greve greve@nmr.mgh.harvard.edu wrote:
Hi Sinead, that all looks good. The only thing I would do differently would be to test whether there is a difference in age-area slopes between TT and TG (contrast 0 0 0 0 -.5 -.5 .5 .5). If there is, then that makes the difference of the intercepts between the areas of the groups. If no difference that survives multiple comparisons, then you can re-run your group difference using the DOSS model, ie, add "doss" after the fsgd file on the mri_glmfit command line, and use a contrast of [-.5 -.5 .5 .5 0]. The DOSS model forces the slopes to be the same which assures that the difference in intercept is interpretable.
Finally, if you are using version 5.1 or lower, you will need to get a patch and rerun mris_preproc.
For a description of the method used for correction, you can use the text from Greve, et al, A Surface-based Analysis of Language Lateralization and Cortical Asymmetry.http://www.researchgate.net/publication/236906359_A_Surface-based_Analysis_of_Language_Lateralization_and_Cortical_AsymmetryJournal of Cog Neurosci. 2013.
doug
On 7/30/13 7:53 AM, Sinead Kelly wrote:
Dear freesurfer users,
I would really appreciate if anybody could confirm that I have preformed my analysis correctly as I have very little experience with freesurfer and I want to make sure that my result is a true finding. Firstly, I am looking at the effect of a schizophrenia risk variant on cortical surface area in a large sample of healthy controls.
a. I performed preprocessing on all of my subjects using *recon-all.* b. I then qcached my data (resampling it to the average subject -*fsaverage *) c. An FSGD file was then created for statistical analysis to look at differences between TT carriers and TG carriers with age and gender as covariates of no interest. Here is an example of the file I created:
GroupDescriptorFile 1 Title MIR137 Class TTMale Class TTFemale Class TGMale Class TGFemale
Variables Age Input CON3140 TGFemale 37 Input CON3891 TTFemale 33 Input CON4664 TTFemale 40 Input CON5591 TTMale 41 Input CON7009 TGMale 21 Input CON7024 TTMale 59 Input CON7038 TTMale 56 Input CON7050 TTMale 30 Input CON7054 TTMale 43 Input CON7055 TTMale 43 Input CON7072 TTFemale 41
d. I applied the following contrast:
-0.5 -0.5 0.5 0.5 0 0 0 0
e. I followed the online tutorial by firstly running the *mris_preproc*script with *area.fwhm10.fsaverage* as my input file. Is this correct? (I ran this separately for the left and right hemisphere). f. I then ran the *mri_glmfi*t script followed by the *mri_glmfit-sim*script: *mri_glmfit-sim --glmdir rh.MIR137_area.glmdir --sim mc-z 5000 1.3 mc-z.abs --sim-sign abs --cwpvalthresh .999 --overwrite* (I ran this separately for each hemisphere)
Does this look correct?
When this ran I then checked the *mc-z.abs.sig.cluster.summary file:* ClusterNo Max VtxMax Size(mm^2) MNIX MNIY MNIZ CWP CWPLow CWPHi NVtxs Annot 1 -3.298 36614 541.01 -26.0 -53.1 62.2 0.77340 0.76580 0.78100 1177 superiorparietal 2 2.918 9004 1642.43 -8.5 9.7 65.2 *0.01020* 0.00840 0.01200 2790 superiorfrontal 3 2.626 126211 363.55 -30.3 -65.2 40.5 0.98580 0.98360 0.98800 817 inferiorparietal 4 2.441 6490 313.02 -30.5 -35.1 58.4 0.99660 0.99560 0.99760 693 postcentral 5 2.375 107140 1095.56 -10.9 -57.2 44.7 0.09720 0.09180 0.10260 2308 precuneus 6 2.347 132687 414.79 -38.6 2.1 24.8 0.95300 0.94920 0.95680 974 precentral 7 2.284 97632 1510.86 -41.4 -50.0 -14.4 * 0.01820* 0.01580 0.02060 2446 fusiform 8 2.122 111909 365.94 -49.2 -29.5 -3.0 0.98480 0.98260 0.98700 837 superiortemporal 9 2.070 74111 444.95 -54.7 -42.0 6.3 0.92120 0.91620 0.92600 998 bankssts 10 1.809 85712 281.07 -57.0 -30.8 21.7 0.99860 0.99800 0.99920 598 supramarginal 11 1.796 121724 312.87 -10.0 64.3 6.5 0.99660 0.99560 0.99760 485 superiorfrontal
I have two significant results here - does this look okay? Does this result indicate that surface area of the fusiform and superior frontal significantly differ between TT and TG carriers (corrected for multiple comparisons?).
How can I check or report the statistics? Is it possible to extract the values and report descriptive as well as inferential statistics for the result?
I would really appreciate if anyone can validate this for me as I have very little experience with freesurfer.
Thank you for your help!
Sinead
Sinead Kelly Neuropsychiatric Genetics Group Trinity Centre St. James's Hospital Dublin 8
Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://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
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.