S.O.S diffusion guys, poor and highly stressed Computer Science grad student here, trying to break into the neuroscience community! :)
Matt, in case this helps, here are the commands I used to go from the released ROI files to the subject ROIs, in order: 1) Create .label.gii files from both left and right hemisphere ROI files. wb_command -cifti-separate Q1-Q6_RelatedParcellation210.L.CorticalAreas_dil_Final_Final_Areas_Group.32k_fs_LR.dlabel.nii COLUMN -label CORTEX_LEFT LEFT.label.gii wb_command -cifti-separate Q1-Q6_RelatedParcellation210.R.CorticalAreas_dil_Final_Final_Areas_Group.32k_fs_LR.dlabel.nii COLUMN -label CORTEX_RIGHT RIGHT.label.gii 2) Convert the .label.gii files to .func.gii files. wb_command -gifti-all-labels-to-rois LEFT.label.gii 1 LEFT.func.gii wb_command -gifti-all-labels-to-rois RIGHT.label.gii 1 RIGHT.func.gii 3) Break each .func.gii file into 180 .func.gii files, one for each ROI. wb_command -metric-merge left-roi-1.func.gii -metric LEFT.func.gii -column 1 (the above command is run 180 times, with -column ranging from 1 to 180 and the output file name being left-roi-<i>.func.gii where <i> ranges from 1 to 180; I repeated the same thing another 180 times to obtain right-ROIs as well) 4) Convert each .func.gii ROI file into subject .gii ROI file using surf2surf surf2surf -i <surface_file> -o left-roi-1.gii --values=left-roi-1.func.gii Here, <surface_file> is <subject_number>.L.white_MSMAll.32k_fs_LR.surf.gii for a left-hemisphere ROI and <subject_number>.R.white_MSMAll.32k_fs_LR.surf.gii for a right-hemisphere ROI. Now these .gii files are the ones I provide to probtrackx. Regards, Karthik ________________________________ From: Glasser, Matthew <glass...@wustl.edu> Sent: Friday, September 1, 2017 4:30:31 PM To: Gopalakrishnan, Karthik; hcp-users@humanconnectome.org Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation I don’t immediately see an issue with your calls to probtrackx, perhaps one of the diffusion guys will take a look and find something… Matt. From: "Gopalakrishnan, Karthik" <gkart...@gatech.edu<mailto:gkart...@gatech.edu>> Date: Friday, September 1, 2017 at 3:27 PM To: Matt Glasser <glass...@wustl.edu<mailto:glass...@wustl.edu>>, "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation Yes, both surface files I mentioned previously are the ones that are provided in the following directory: ${Subject}/T1w/fsaverage_LR32k Thanks! Karthik ________________________________ From: Glasser, Matthew <glass...@wustl.edu<mailto:glass...@wustl.edu>> Sent: Friday, September 1, 2017 4:21:43 PM To: Gopalakrishnan, Karthik; hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation Are all of these files in ${StudyFolder}/${Subject}/T1w? Matt. From: "Gopalakrishnan, Karthik" <gkart...@gatech.edu<mailto:gkart...@gatech.edu>> Date: Friday, September 1, 2017 at 3:19 PM To: Matt Glasser <glass...@wustl.edu<mailto:glass...@wustl.edu>>, "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation Yes, I used the released bedpostX data. The white matter surfaces used were: 100206.L.white_MSMAll.32k_fs_LR.surf.gii 100206.R.white_MSMAll.32k_fs_LR.surf.gii Thanks! Karthik ________________________________ From: Glasser, Matthew <glass...@wustl.edu<mailto:glass...@wustl.edu>> Sent: Friday, September 1, 2017 4:14:03 PM To: Gopalakrishnan, Karthik; hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation Which white matter surface are those on? I assume the bedpostX data are what we released? Peace, Matt. From: "Gopalakrishnan, Karthik" <gkart...@gatech.edu<mailto:gkart...@gatech.edu>> Date: Friday, September 1, 2017 at 3:11 PM To: Matt Glasser <glass...@wustl.edu<mailto:glass...@wustl.edu>>, "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation The file white-masks.txt and the file targets.txt contain newline-separated paths to each of the 360 ROIs (.gii files). Thanks! Karthik ________________________________ From: Glasser, Matthew <glass...@wustl.edu<mailto:glass...@wustl.edu>> Sent: Friday, September 1, 2017 4:06:15 PM To: Gopalakrishnan, Karthik; hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation What is in the .txt? Peace, Matt. From: "Gopalakrishnan, Karthik" <gkart...@gatech.edu<mailto:gkart...@gatech.edu>> Date: Friday, September 1, 2017 at 3:04 PM To: Matt Glasser <glass...@wustl.edu<mailto:glass...@wustl.edu>>, "hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>" <hcp-users@humanconnectome.org<mailto:hcp-users@humanconnectome.org>> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation Hi Matt, Here's the probtrackx calls: For network mode: probtrackx2 --network -x white-masks.txt -l --onewaycondition -c 0.2 -S 2000 --steplength=0.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --sampvox=0.0 --forcedir --opd -s Diffusion.bedpostX/merged -m Diffusion.bedpostX/nodif_brain_mask.nii.gz --dir=WHITE_NETWORKMODE For seed-to-ROI mode (in this example, the seed ROI is L_1.gii, but I did this for all 360 ROIs): probtrackx2 --seed=../whiteROIs/L_1.gii -l -c 0.3 -S 2000 --steplength=0.3 -P 5000 --fibthresh=0.01 --distthresh=20.0 --forcedir --opd -s ../Diffusion.bedpostX/merged -m ../Diffusion.bedpostX/nodif_brain_mask --dir=L1 --targetmasks=targets.txt --os2t --s2tastext Thanks! Karthik ________________________________ From: Glasser, Matthew <glass...@wustl.edu<mailto:glass...@wustl.edu>> Sent: Friday, September 1, 2017 8:16:51 AM To: Gopalakrishnan, Karthik; HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org> Subject: Re: [HCP-Users] probtrackx network with HCP data and MMP parcellation Can you post your probtrackx call? Peace, Matt. From: <hcp-users-boun...@humanconnectome.org<mailto:hcp-users-boun...@humanconnectome.org>> on behalf of "Gopalakrishnan, Karthik" <gkart...@gatech.edu<mailto:gkart...@gatech.edu>> Date: Thursday, August 31, 2017 at 9:27 PM To: "HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org>" <HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org>> Subject: [HCP-Users] probtrackx network with HCP data and MMP parcellation Hi all, I ran probtrackx for HCP subject 100206 using the MMP1.0 Glasser surface ROIs. My seeds in each ROI are at the white matter-gray matter boundary. I ran probtrackx in network mode as well as seed-to-ROI mode for the subject, and I applied the method described in http://journal.frontiersin.org/article/10.3389/fninf.2016.00046/full to infer structural networks from the resultant data from probtrackx. What I observe is that the networks aren't particularly dense. Since MMP1.0 has 360 ROIs, there are 360*359 = 129240 possible directed edges, but the networks I infer have about 2300 edges, which is a density of ~2%. Could someone share their insights on why I'm observing such low-density networks/where I might be doing something wrong? Thanks! Karthik _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org<mailto:HCP-Users@humanconnectome.org> http://lists.humanconnectome.org/mailman/listinfo/hcp-users _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
