Hi Eric   I don’t see any fundamental problem with this analysis. It looks like a regular lme computation. The p-value is essentially zero (to the extent of numeric precision). Even under the null hypothesis p-values distribute uniform(0,1). When signal is present in the data the distribution of p-values is skewed toward zero.      Best -Jorge

El Jueves 20 de marzo de 2014 23:31, Eric Cunningham etc42@hawaii.edu escribió:

Hello Freesurfer experts,

I am getting some

unusual results with the lme pipeline, and I would appreciate help with identifying the most likely step in which the problem may have originated.

Running lme_fit_FS.m on a single vertex, followed by

lme_F on the output of that (stats) yields a p-value of 0.  This is because the fcdf function in matlab rounds to 1 if the F-value is big enough (in our case it rounds at about 80, and our F-value is 123).

This being the case, we are trying to determine whether this is (1) a problem with the data, (2) a problem with the script somewhere, or (3) a system setting I don't know about.

I'm not sure

if you have enough information, but can you tell me if the F-values and Beta values look possible (reasonable?)?

Thanks for any advice you can provide.

-E

Starting Fisher scoring iterations Likelihood at FS iteration 1 : 484.5267 Gradient norm: 53.7987 Likelihood at FS iteration 2 : 484.5984 Gradient norm: 1.3858 Likelihood at FS iteration 3 : 484.5985 Gradient norm: 0.022222 Likelihood at FS iteration 4 : 484.5985 Gradient norm: 0.00056884 Total elapsed time is 0.24038 seconds

stats =

Bhat: [7x1 double]      CovBhat: [7x7 double]        bihat: [1x93 double]     Covbihat: [93x1 double]     phisqhat: 0.0110        SIGMA: [345x5 double]            W: [345x5 double]         Dhat: 0.0213            X: [345x7 double]        Zcols: 1           re: [345x1 double]           ni: [93x1 double]        lreml: 484.5985

st =

1

K>> stats.Bhat

ans =

3.1531    -0.0559     0.0148    -0.0154    -0.0309    -0.0767    -0.0229

K>>   -0.0559 K>>  fstats = lme_F(stats,contrastmatrix)

fstats =

F: 123.4649     pval: 0      sgn: -1       df: [1 284.0811] _______________________________________________ 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.