3dMVM gltCode error

Hey Gang,

I’m having trouble with my 3dMVM script below. I’m trying to run an ANCOVA with functional connectivity maps of 3 seed ROIs (BNST, BLA, CEN) as repeated measures for each subject. The first 3 glt’s are to look at just functional connectivity of each ROI while controlling for the other 2. Then the next 6 glt’s are to look at how each ROI is related to one of two anxiety measure covariates (STAI and IUS) while controlling for the other.

Here’s my 3dMVM script and data table:


3dMVM -prefix ANCOVA_extendedamy_ANX_IUS -jobs 1 \
          -bsVars "gender+IUS+STAI" \
          -wsVars ROI \
          -qVars "STAI,IUS" \
	  -qVarsCenters "36.98,59.18" \
	  -mask mean_epi_mask+tlrc \
          -num_glt 9 \
          -gltLabel 1 BNST -gltCode 1 'ROI : 1*BNST' \
	-gltLabel 2 BLA -gltCode 2 'ROI : 1*BLA' \
	-gltLabel 3 CEN -gltCode 3 'ROI : 1*CEN' \
	-gltLabel 4 BNST_STAI -gltCode 4 'ROI : 1*BNST STAI' \
	-gltLabel 5 BLA_STAI -gltCode 5 'ROI : 1*BLA STAI' \
	-gltLabel 6 CEN_STAI -gltCode 6 'ROI : 1*CEN STAI' \
	-gltLabel 7 BNST_IUS -gltCode 7 'ROI : 1*BNST IUS' \
	-gltLabel 8 BLA_IUS -gltCode 8 'ROI : 1*BLA IUS' \
	-gltLabel 9 CEN_IUS -gltCode 9 'ROI : 1*CEN IUS' \
	-dataTable @3dMVM_Table.txt

Here’s my dataTable (3dMVM_Table.txt) for 4 of my 54 subjects:


Subj	gender	IUS	STAI	ROI	InputFile	
s6301	2	78	29	BNST	'/raid-09/LS/Data/7T_fall2017/6301/rest/BNST_r2z+tlrc'[0]''	\
s6301	2	78	29	BLA	        '/raid-09/LS/Data/7T_fall2017/6301/rest/AMY_BLA_r2z+tlrc'[0]''	\
s6301	2	78	29	CEN	'/raid-09/LS/Data/7T_fall2017/6301/rest/AMY_CEN_r2z+tlrc'[0]''	\
s6302	2	43	31	BNST	'/raid-09/LS/Data/7T_fall2017/6302/rest/BNST_r2z+tlrc'[0]''	\
s6302	2	43	31	BLA	        '/raid-09/LS/Data/7T_fall2017/6302/rest/AMY_BLA_r2z+tlrc'[0]''	\
s6302	2	43	31	CEN	'/raid-09/LS/Data/7T_fall2017/6302/rest/AMY_CEN_r2z+tlrc'[0]''	\
s6303	2	78	37	BNST	'/raid-09/LS/Data/7T_fall2017/6303/rest/BNST_r2z+tlrc'[0]''	\
s6303	2	78	37	BLA	        '/raid-09/LS/Data/7T_fall2017/6303/rest/AMY_BLA_r2z+tlrc'[0]''	\
s6303	2	78	37	CEN	'/raid-09/LS/Data/7T_fall2017/6303/rest/AMY_CEN_r2z+tlrc'[0]''	\
s6304	1	51	32	BNST	'/raid-09/LS/Data/7T_fall2017/6304/rest/BNST_r2z+tlrc'[0]''	\
s6304	1	51	32	BLA	        '/raid-09/LS/Data/7T_fall2017/6304/rest/AMY_BLA_r2z+tlrc'[0]''	\
s6304	1	51	32	CEN	'/raid-09/LS/Data/7T_fall2017/6304/rest/AMY_CEN_r2z+tlrc'[0]''	\
.
.
.

First, I want to make sure those gltCodes are set up correctly, but also I keep getting an error that pertains to specifically the BLA and CEN ROI variables. When running the script above, I get the following error:


**Error: Inappropriate coding in test No.2! 
Warning messages: 
1: In process.MVM.opts(lop, verb=lop$verb) : NAs introduced by coercion
2: In process.MVM.opts(lop, verb=lop$verb) : NAs introduced by coercion

So it seems to be making it past the first glt that looks at BNST as the ROI, and trips on the second glt where BLA is the ROI. I’ve rearranged the order of these glt’s and if I put the BLA or the CEN glt first then it’ll error out at the 1st glt test. I’ve triple checked spelling and spacing and my data table and cannot understand what the issue is with BLA or CEN and not with BNST. Is this an issue with how the ROI variable is getting parsed from my dataTable?

Thanks in advance!

Carissa

Carissa,

I don’t see anything suspicious with the second or third test, but the following tests

-gltLabel 4 BNST_STAI -gltCode 4 'ROI : 1*BNST STAI' \
-gltLabel 5 BLA_STAI -gltCode 5 'ROI : 1*BLA STAI' \
-gltLabel 6 CEN_STAI -gltCode 6 'ROI : 1*CEN STAI' \
-gltLabel 7 BNST_IUS -gltCode 7 'ROI : 1*BNST IUS' \
-gltLabel 8 BLA_IUS -gltCode 8 'ROI : 1*BLA IUS' \
-gltLabel 9 CEN_IUS -gltCode 9 'ROI : 1*CEN IUS' \

should be

-gltLabel 4 BNST_STAI -gltCode 4 'ROI : 1*BNST STAI :' \
-gltLabel 5 BLA_STAI -gltCode 5 'ROI : 1*BLA STAI :' \
-gltLabel 6 CEN_STAI -gltCode 6 'ROI : 1*CEN STAI :' \
-gltLabel 7 BNST_IUS -gltCode 7 'ROI : 1*BNST IUS :' \
-gltLabel 8 BLA_IUS -gltCode 8 'ROI : 1*BLA IUS :' \
-gltLabel 9 CEN_IUS -gltCode 9 'ROI : 1*CEN IUS :' \