Hi Marco, I'm copying the list on this because this is a good question about an aspect that is not well documented.
The answer to the question has to do with the sign of the test. By default, mri_glmfit produces p-values that are for unsigned tests. The p-value for a signed test if half that of the signed test (ie, it is more significant). When the p-value is expressed as -log10(p), then the signed test will be 0.301 higher than the unsigned test. Eg, if you have a p-value of .01, the value in the sig.mgh file will be 2.0, but for a signed test it would be .005, making the value in the sig.mgh file 2.3. When you run mri_glmfit-sim and specify a signed test (ie, --sim-sign pos or --sim-sign neg), it lowers the threshold to account for this. So below you specified 1.3 as the threshold and --sim-sign pos, so it lowered the threshold to 1.0. hope this is clear doug Marco Loggia, PhD wrote: > Sorry Doug, > > Below are directory and command line. Also attached is a screenshot showing > a vertex located on a significant cluster, but below threshold in the > sig.mgh. > > Thanks! > Marco > > dir=/autofs/cluster/randy/2/lbp/freesurfer/patients+controls/analyses/lh.lbp > +ctrl.session1.asl2-asl1_30sept10_sm7_03mm_masked_projfrac-max_P=0.05.glmdir > /lbp_only_24sept10 > > > pbsubmit -c "mri_glmfit-sim --glmdir > lh.lbp+ctrl.session1.asl2-asl1_30sept10_sm7_03mm_masked_projfrac-max_P=0.05. > glmdir > --sim perm 10000 1.30103 perm.positive --sim-sign pos --overwrite" > > > > -----Original Message----- > From: Douglas N Greve [mailto:gr...@nmr.mgh.harvard.edu] > Sent: Monday, October 18, 2010 10:34 AM > To: Marco Loggia, PhD > Subject: Re: montecarlo > > Hi Marco, can you send the command-line and directory that you ran it > from? In general, it is always a good idea to send this ... > > doug > > Marco Loggia, PhD wrote: > >> Hello Doug, >> >> >> >> I have one question about the montecarlo simulation. >> >> >> >> When I compare the mc-z.positive.sig.cluster.mgh and the >> mc-z.positive.sig.masked.mgh, I notice that the significant clusters >> include also some vertices that are below the cluster forming >> threshold. Is this normal? Why is that? >> >> >> >> Thanks, >> >> >> >> Marco >> >> >> _____________________ >> >> >> >> Marco L. Loggia, PhD >> >> Postdoctoral Fellow, Mass General Hospital & Brigham and Women's Hospital >> >> Harvard Medical School >> >> >> >> CNY Building 120, suite 101E >> >> Charlestown, Massachusetts 02129 >> >> Phone: (617) 643-7267 >> >> Fax: (617) 643-7340 ma...@nmr.mgh.harvard.edu >> >> >> >> > > > > ------------------------------------------------------------------------ > -- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html _______________________________________________ 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.
