Extending (Smith et al., 2015) to full HCP dataset

This repo is based on the Arxiv preprint Computationally replicating the Smith et al. (2015) positive-negative mode linking functional connectivity and subject measures. The authors' github repo is outdated but they provide a link to OSF. This repo is based on the OSF project of this preprint.

The files you need (in data):

• RESTRICTED.csv: the "restricted data" from ConnectomeDB, originally named like RESTRICTED_yourname_MM_DD_YY_HH_MM_SS.csv. Requires signing the restricted data use term.
• UNRESTRICTED.csv: the "behavioral data" from ConnectomeDB, originally named like UNRESTRICTED_yourname_MM_DD_YY_HH_MM_SS.csv.
• netmats2.txt: the connectivity matrix from ConnectomDB. It is in the folder netmats/3T_HCP1200_MSMALL_d200_ts2 after unzipping the file netmats_3T_HCP1200_MSMAll_ICAd200_ts2.tar.gz in the HCP_PTN1200 folder of the downloaded zip file HCP1200_Parcellation_Timeseries_Netmats.zip.
• subject_list.txt (provided here): ID of the subjects in the netmat file. Originally called subjectIDs.txt. Also in the HCP_PTN1200 folder.
• column_headers.txt (provided here): names of the subject measures, obtained from HCP-CCA website but changing the first variable from 'Subject ID' to 'Subject'.

To run the scripts:

1. Switch to scripts.
2. Run the matlab script summarize_movement.m to generate a HCP1200-Motion-Summaries-rfMRI.csv under the data folder. You'll need to modify the 'HCP_dir' in the first line of code.
3. Run the python script NET.py by python3 NET.py ../data/netmats2.txt 200 1003 (netmats location, #ICA components, #subjects) to generate a NET.txt under processed_data.
4. Run the python script sm_file_update.py by python3 sm_file_update.py ../data/UNRESTRICTED.csv ../data/RESTRICTED.csv ../data/subject_list.txt ../processed_data/ to generate b1200_m.csv (modified unrestricted "behavioral" data), r1200_m.csv (modified restricted data), hcp1200_family_data.csv (family info) under processed_data.
5. Run the python script VARS.py to generate VARS.txt under processed_data.
6. Run the matlab script hcp_1200_cca_final.m to do the CCA.