Release notes for HCP_S1200_GroupAvg_v1.zip 
31 July 2017
This open access dataset has been made freely available to the research community by the WU-Minn HCP consortium (http://www.humanconnectome.org/).  This document contains a brief overview plus links to additional information.

1) Open Access Data Use Terms:  Before using this dataset you must agree to the HCP Open Access Data Use Terms at http://humanconnectome.org/data/data-use-terms/DataUseTerms-HCP-Open-Access-26Apr2013.pdf 

2) What's in the dataset:   This dataset includes group-average structural and functional MRI data for the HCP S1200 data release (March 2017). The S1200 dataset includes both cortical areal feature-based surface registration (MSMAll, using folding, myelin maps, and rfMRI) and the purely folding-based registration (MSMSulc) used for the S500 and earlier HCP data releases; MSM stands for Multimodal Surface Matching (Robinson et al., Neuroimage, 2014).  Also included in this dataset are S1200.All* composite files containing individual MSMAll-registered maps of folding, sulc, myelin, and thickness for all S1200 subjects with structural imaging data. The individual subject data may be scrolled through in wb_view efficiently as separate maps within each of these files. 

The group average data for the S1200 release involve modestly different numbers of HCP subjects for the different modalities. Group average structural data are based on 1096 subjects that have at least one run of rfMRI data (and hence were eligible for MSMAll registration). Functional connectivity is based on the 1003 subjects having complete (4 x 15min) rfMRI runs, or a subset of 812 of these subjects whose fMRI data was reconstructed using the improved r227 recon algorithm that was in place in the latter two thirds of the HCP project. Task fMRI data is based on 997 subjects that additionally have complete tfMRI data for all 7 tasks (r177+r227 recon versions). These groups all contain some subjects that are related to some members of the group, thus they are referred to as R1096, R1003, R812, and R997 (R designating related).

The task-fMRI and functional connectivity data have been processed and additionally analyzed as described in the Group-average functional data section of the HCP 1200 Subjects Release Reference Manual. The accompanying scene file uses surfaces and volumes generated from the R1096 MSMAll registered group-average as a structural atlas to display and compare the group average functional data in Workbench.

3) How to view the data:  The dataset is customized for viewing with the wb_view visualization platform in Connectome Workbench v1.2.3 software (http://www.humanconnectome.org/connectome/get-connectome-workbench.html). Some of the files are in standard NIFTI and GIFTI format and can also be viewed using other brain-mapping software platforms.  Many files are in CIFTI-2 file format (http://www.nitrc.org/projects/cifti/), which can incorporate surface vertices (both hemispheres) plus subcortical gray-matter voxels into a single *.dscalar.nii file format used for Task-fMRI data. Other CIFTI-2 files (e.g., cortical myelin maps and thickness maps) include data from both left and right cerebral  hemispheres in a single file (but without subcortical voxels). The scene file included (HCP_S1200_GroupAvg_v1.scene) enables rapid access to conveniently formatted displays for each modality.

4) For guidance on navigating wb_view and visualizing the data, we recommend the tutorial document ConnectomeWorkbenchTutorial_HCP_S1200_GroupAvg_v1.pdf included in this zip archive .  It includes instructions for navigating 5 'scenes' that enable quick visualization of each modality. 

5) Access to the R1003 and R812 group-average functional connectivity maps in Connectome Workbench.  The S1200 data release includes two MSMAll registered functional connectivity matrices ('dense functional connectomes) that are accessed interactively via the ConnectomeDB database (see the tutorial document), one based on all R1003 subjects with complete fMRI data and another based on the R812 subjects with complete data AND r227 image reconstruction. These maps are highly similar, see a comparison between the R1003 and R812 rfMRI data as an appendix to the S1200 high-level connectivity analysis documentation available at https://www.humanconnectome.org/documentation/S1200/HCP1200-DenseConnectome+PTN-July2017.pdf

6) Download of group-average functional connectivity matrix files. The group-average "dense" functional connectome files for the R820 group are also available for download as a separate package (due to their large size, 33GB per file) via the ConnectomeDB HCP project page (https://db.humanconnectome.org/data/projects/HCP_1200): 

HCP_S1200_1003_rfMRI_MSMAll_groupPCA_d4500ROW_zcorr.dconn.nii
HCP_S1200_812_rfMRI_MSMAll_groupPCA_d4500ROW_zcorr_recon2.dconn.nii

Also available on the ConnectomeDB HCP project page are the corresponding dense timeseries files:
HCP_S1200_1003_rfMRI_MSMAll_groupPCA_d4500_Eigenmaps.dtseries.nii
HCP_S1200_812_rfMRI_MSMAll_groupPCA_d4500_Eigenmaps_recon2.dtseries.nii
which are smaller files (2GB) and can be used as input to group-ICA and to create the dense connectomes above with scripts included in the Parcellations, Timeseries, Netmats (PTN) dataset. 

7) The cortical parcellations included are the 180-area (per hemisphere) multimodal parcellation HCP_MMP1.0 (Q1-Q6_RelatedValidation210.CorticalAreas_dil_Final_Final_Areas_Group_Colors.32k_fs_LR.dlabel.nii) from Glasser et al. (2016), along with parcellations from Gordon et al. (2016), Power et al. (2014), Yeo et al. (2011), and Van Essen et al. (2012).  Accompanying the Gordon et al. parcellation is a key (Gordon333_Key.txt) that maps parcel numbers to those in a (differently ordered) dlabel.nii file accessible at http://www.nil.wustl.edu/labs/petersen/Resources.html.  

8) Additional information on the data acquisition and analysis methods for this data are available in the HCP S1200 Data Release Reference Manual (http://www.humanconnectome.org/documentation/S1200) and in several articles in a special issue of NeuroImage, including Barch et al., (2013) and Glasser et al., (2013), available at http://www.sciencedirect.com/science/article/pii/S1053811913005272 and http://www.sciencedirect.com/science/article/pii/S1053811913005053, respectively). 
 
9) Acknowledge the funding source. If you make use of these data in a publication or public presentation, you should include the following acknowledgment: "Data were provided [in part] by the Human Connectome Project, WU-Minn Consortium (Principal Investigators: David Van Essen and Kamil Ugurbil; 1U54MH091657) funded by the 16 NIH Institutes and Centers that support the NIH Blueprint for Neuroscience Research; and by the McDonnell Center for Systems Neuroscience at Washington University."

10) Cite relevant publications.  See http://www.humanconnectome.org/study/hcp-young-adult/document/hcp-citations for additional information on how to cite HCP publications if you have used data provided by the WU-Minn HCP consortium.
