colelab/coleanticevicnetpartition

Public release of The Cole-Anticevic Brain-wide Network Partition (CAB-NP)

The Cole-Anticevic Brain-wide Network Partition (CAB-NP)

Authors

Jie Lisa Ji, Marjolein Spronk, Kaustubh Kulkarni, Grega Repovs, Alan Anticevic, and Michael W. Cole

Cole Neurocognition Lab, http://www.colelab.org/

Anticevic Lab, http://anticeviclab.yale.edu/

Version Info and Acknowledgements

Version 1.1.6: Feb 7, 2021, fixed missing values in CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt, removed reference files (Q1-Q6_RelatedValidation210.*.CorticalAreas_dil_Final_Final_Areas_Group_Colors.32k_fs_LR) from original Glasser 2016 parcellation.

Version 1.1.5: Aug 17, 2020, updated CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt and CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_LabelKey.txt to include mapping to original cortical parcel label names from Glasser et al. 2016.

Version 1.1.4: June 11, 2020, updated displayed borders in scene files to render as lines (instead of spheres) to speed up loading time in wb_view.

Version 1.1.3: June 1, 2020, updated all mentions of the Glasser et al. (2016) Multi-modal Parcellation (MMP) to be consistent. The latest "Q1-Q6_RelatedValidation210" version was used.

Version 1.1.2: May 28, 2020, fixed labelling in CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii to be consistent with v1.1 format. Updated + fixed errors in CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt and switched to more intuitive ordering of NetworkSortedOrder. Also added CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dscalar.nii and code to show how ReorderingByNetworks was computed. Analogous updates for woGSR files in NoGSRSubcortex folder.

Version 1.1.1: May 1, 2020, added Glasser360Indices_LR.dscalar.nii. This file can be loaded into Connectome Workbench, such that clicking a parcel will indicate which parcel number it is within the cortical Glasser parcellation. Tip: Turn off Display Borders to speed up Connectome Workbench.

Version 1.1.0: July 8, 2019, updated all files with cortical parcels from the HCP1200 release; updated and standardized naming of parcels; added LabelKey.txt showing mappings for the wSubcorGSR and woSubcorGSR parcellations between: Index; KeyValue; Label; Red; Green; Blue; Alpha; Hemisphere; Network; NetworkKey; NetworkSortedOrder

Version 1.0.5: October 10, 2018, first public release.

Version 1.0.4: October 9, 2018, added cortex+subcortex parcel order files and dscalar versions of CIFTI files.

Version 1.0.3: October 7, 2018, filenames changed for simplicity and consistency.

Version 1.0.2: October 4, 2018, changed subcortical parcellation with global signal regression (GSR) applied as a preprocessing to be the primary version. This is based on results reported in the final version of the article accepted for publication after peer review.

Version 1.0.1: January 17, 2018, added a version of the subcortical parcellation with GSR applied as a preprocess step. This was included due to concern over extensive assignment of subcortical voxels to the visual networks (which GSR reduced). A version of the parcellation based on the conjunction of the GSR and non-GSR versions is included, for those who wish to only use subcortical voxels with assignments consistent with and without GSR.

Version 1.0.0: September 27, 2017

Cite as: Ji JL*, Spronk M*, Kulkarni K, Repovs G, Anticevic A**, Cole MW** (2019). "Mapping the human brain's cortical-subcortical functional network organization". NeuroImage. 185:35–57. doi:10.1016/j.neuroimage.2018.10.006 [* = equal contribution; ** = senior authors] https://doi.org/10.1016/j.neuroimage.2018.10.006 and https://github.com/ColeLab/ColeAnticevicNetPartition/

Interactive versions of the figures from the paper are available here: https://balsa.wustl.edu/study/show/wZML

Scientific article also available as an open access bioRxiv preprint: http://doi.org/10.1101/206292

Data were provided 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. This work was supported by the NIH via awards K99/R00-MH096801 (Cole), DP5-OD012109 (Anticevic), R01-MH109520 (Cole), R01-MH108590 (Anticevic), R01-AG055556 (Cole), and R01-MH112189 (Anticevic), as well as the Brain and Behavior Foundation (NARSAD) Independent Investigator grant (Anticevic) and ARRS J7-6829 (Repovs).

Overview

This network partition was created using the Glasser 2016 parcels [Glasser MF, Coalson TS, Robinson EC, Hacker CD, Harwell J, Yacoub E, Ugurbil K, Andersson J, Beckmann CF, Jenkinson M, Smith SM, Van Essen DC (2016) A multi-modal parcellation of human cerebral cortex. Nature. http://doi.org/10.1038/nature18933], along with data of 337 unrelated healthy volunteers from the WashU-Minn Human Connectome Project (HCP) [https://www.humanconnectome.org/]. Resting-state fMRI data were used (across all 4 resting-state fMRI runs per subject). ICA+FIX was used for denoising and MSMAll for registration (along with standard HCP minimal preprocessing). Pearson correlations between time series were then calculated between all parcels/regions, and network communities were identified using the general Louvain algorithm.

The cortical network partition was extended into subcortex. This was accomplished by labeling each subcortical voxel with the cortical network with the strongest average Pearson correlation.

See the main publication reporting this partition for more information: Ji JL*, Spronk M*, Kulkarni K, Repovs G, Anticevic A**, Cole MW** (2019). "Mapping the human brain's cortical-subcortical functional network organization". NeuroImage. 185:35–57. doi:10.1016/j.neuroimage.2018.10.006 [* = equal contribution; ** = senior authors] https://doi.org/10.1016/j.neuroimage.2018.10.006 [Open access preprint: http://doi.org/10.1101/206292]

The Glasser2016 parcels are available here: https://balsa.wustl.edu/study/show/RVVG

More info on the cortical parcels is available in the main and supplemental documents released as part of the main Glasser2016 publication (e.g., details in Table 1 of the Supplementary Neuroanatomical Results document, page 81): Glasser, Matthew F, Timothy S Coalson, Emma C Robinson, Carl D Hacker, John Harwell, Essa Yacoub, Kamil Ugurbil, et al. (2016). "A Multi-Modal Parcellation of Human Cerebral Cortex." Nature, July. doi:10.1038/nature18933. http://www.nature.com/doifinder/10.1038/nature18933

Evidence that the CIFTI format used here (combining cortical surface with subcortical volume data) is superior to prior methods: Coalson, T.S., Van Essen, D.C., Glasser, M.F., 2018. The impact of traditional neuroimaging methods on the spatial localization of cortical areas. Proc Natl Acad Sci USA 115, E6356–E6365. https://doi.org/10.1073/pnas.1801582115

Software versions this release was tested on: Connectome Workbench 1.3.2, MATLAB R2014b, and Python 2.7

Many features are available in Connectome Workbench's wb_command (https://www.humanconnectome.org/software/workbench-command) for interfacing with CAB-NP.

Quick Start

Download and open Connectome Workbench (https://www.humanconnectome.org/software/connectome-workbench). Load the ColeAnticevicNetPartition.wb.spec file. Click the "Load Scenes" button, select one of the scenes of interest, and click the "Show" button. This will allow you to view and interact with the network partition and parcels.

The Network Partition

Top Left: Illustration of the network partition with the Glasser parcels. The colors correspond to the colors labeled in the network matrix (to the right). Top Right: Axial slices illustrating the subcortical extension of the cortical network partition. Each voxel was assigned to the network that it had the highest mean resting-state functional connectivity with. See Ji et al. (In Press) for more information. Bottom Left: Network matrix with Pearson correlation-based resting-state functional connectivity, sorted based on community affiliation according to the network partition. An fMRI dataset of 337 subjects from the WashU-Minn Human Connectome Project (HCP) was used (https://www.humanconnectome.org/), with 4 runs for each subject. See Ji et al. (In Press) for more information. Bottom Right: Same as bottom left, but now also including subcortical parcels.

The partition across transaxial slices of the S1200 HCP average T1 image.

Included Files

• ColeAnticevicNetPartition.wb.spec - Main Connectome Workbench file, specifying network partition visualization files. Load this file with wb_view to visualize and interact with the network partition.
• ColeAnticevicNetPartition_MainScene.wb.scene - The main Connectome Workbench scene file.
• ColeAnticevicNetPartition_OtherScenes.wb.scene - Alternative scenes for use with Connectome Workbench. These were separated from the main scene file to accommodate computers with low amounts of RAM.
• cortex_community_order.mat - The order the Glasser parcels should be in to reveal the community structure identified by this network partition, in MATLAB format. Note that this file assumes you have the left hemisphere Glasser parcellation regions first, followed by the right hemisphere regions.
• cortex_community_order.txt - Same as the previous file, but in text format.
• cortex_subcortex_community_order.mat - The order of all 718 regions (cortex+subcortex), sorted by network affiliation, in MATLAB format.
• cortex_subcortex_community_order.txt - The order of all 718 regions (cortex+subcortex), sorted by network affiliation, in plain text format.
• cortex_subcortex_parcel_network_assignments.mat - A vector of numbers, one per cortex+subcortex parcel, indicating which network that parcel was assigned to in the network partition (in MATLAB format). (Parcel order: L first, R second.)
• cortex_subcortex_parcel_network_assignments.txt - Same as the previous file, but in text format.
• LoadParcellatedDataInMatlab_Example.m - Example of how to parcellate CIFTI fMRI data and load it into MATLAB
• LoadParcellatedDataInMatlab_Example_cortexonly.m - Example of how to parcellate CIFTI fMRI data and load it into MATLAB, using cortical parcels only (no subcortical parcels)
• LoadParcellatedDataInPython_Example.py - Example of how to parcellate CIFTI fMRI data and load it into Python
• LoadParcellatedDataInPython_Example_cortexonly.py - Example of how to parcellate CIFTI fMRI data and load it into Python, using cortical parcels only (no subcortical parcels)
• network_labelfile.txt - The labels for each network, along with color information (RGBA value).
• cortex_parcel_network_assignments.mat - A vector of numbers, one per cortical parcel, indicating which network that parcel was assigned to in the network partition (in MATLAB format). (Parcel order: L first, R second.)
• cortex_parcel_network_assignments.txt - Same as the previous file, but in text format.
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_netassignments_LR.dlabel.nii - Cortex + subcortex (whole-brain) network assignments. Global signal regression (GSR) applied to subcortical voxels as a preprocessing step.
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii - Same as above, but at the parcel level (rather than the network assignment level).
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_v1.1_LR.txt - Label text file for CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii
• Glasser360Indices_LR.dscalar.nii - Indicates the parcel number (ordered left hemisphere, then right hemisphere) for each cortical parcel. This is especially useful when loaded into Connectome Workbench along with the main scene file, since simply clicking on a parcel will indicate the parcel number for that parcel. Tip: Turn off Display Borders to speed up Connectome Workbench.
• Q1-Q6_RelatedValidation210.L.CorticalAreas_dil_Final_Final_Areas_Group.32k_fs_LR.border - The latest version of the left-hemisphere borders of the Glasser 2016 parcels
• Q1-Q6_RelatedValidation210.R.CorticalAreas_dil_Final_Final_Areas_Group.32k_fs_LR.border - The latest version of the right-hemisphere borders of the Glasser 2016 parcels
• README.md - This file
• S1200_AverageT1w_restore.nii.gz - The average of 1096 subjects from the HCP dataset, from the S1200 release. From HCP_S1200_GroupAvg_v1.zip. For more info visit http://www.humanconnectome.org/documentation/S1200 and https://www.humanconnectome.org/study/hcp-young-adult/article/s1200-group-average-data-release. Before using data from HCP 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
• S1200.L.inflated_MSMAll.32k_fs_LR.surf.gii - Left hemisphere inflated cortical surface
• S1200.L.midthickness_MSMAll.32k_fs_LR.surf.gii - Left hemisphere midthickness cortical surface
• S1200.L.pial_MSMAll.32k_fs_LR.surf.gii - Left hemisphere pial cortical surface
• S1200.L.very_inflated_MSMAll.32k_fs_LR.surf.gii - Left hemisphere very inflated cortical surface
• S1200.R.inflated_MSMAll.32k_fs_LR.surf.gii - Right hemisphere inflated cortical surface
• S1200.R.midthickness_MSMAll.32k_fs_LR.surf.gii - Right hemisphere midthickness cortical surface
• S1200.R.pial_MSMAll.32k_fs_LR.surf.gii - Right hemisphere pial cortical surface
• S1200.R.very_inflated_MSMAll.32k_fs_LR.surf.gii - Right hemisphere very inflated cortical surface
• S1200.sulc_MSMAll.32k_fs_LR.dscalar.nii - Cortical surface sulcus pattern for visualization of cortical surface
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii - The parcels (same as in CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii) sorted by network according to cortex_subcortex_parcel_network_assignments.txt. The order of parcels within a network is sorted as such: Cortex then Subcortex; L Hemisphere then R Hemisphere then LR.
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dscalar.nii - The dscalar file corresponding to CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii .
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_v1.1_LR_ReorderedByNetworks.txt - - The labels for CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii, along with color information (RGBA value).
• CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt - Legend text file showing all mappings for the wSubcorGSR parcellation between: Index; KeyValue; Label; Red; Green; Blue; Alpha; Hemisphere; Network; NetworkKey; NetworkSortedOrder, GlasserLabelName
• CortexSubcortex_ColeAnticevic_NetPartition-ReorderedbyNetworks.sh - The code used to sort parcels by network to create CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii and CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii.
• CAB-NP_v1.1_Labels-ReorderedbyNetworks.xlsx - Excel sheet used to sort parcels by network to create CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii and CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii.
• SeparateHemispheres directory - Files with each hemisphere separated. This can be useful for ensuring that parcels from each hemisphere are in the correct order by loading each hemisphere separately (e.g., in MATLAB).
  • cortex_ColeAnticevic_NetPartition_GlasserParcels_L.label.gii - Left hemisphere cortex-only partition
  • cortex_ColeAnticevic_NetPartition_GlasserParcels_R.label.gii - Right hemisphere cortex-only partition
  • CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_netassignments_L.dlabel.nii - Left hemisphere cortex+subcortex network assignments for each parcel
  • CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_netassignments_R.dlabel.nii - Right hemisphere cortex+subcortex network assignments for each parcel
  • CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_L.dlabel.nii - Left hemisphere cortex+subcortex list of parcels, with a label for each parcel. Note that some midline subcortical parcels were split to create this left-hemisphere-only version, such that combining both hemispheres results in 758 parcels (rather than 718).
  • CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_R.dlabel.nii - Right hemisphere cortex+subcortex list of parcels, with a label for each parcel. Note that some midline subcortical parcels were split to create this right-hemisphere-only version, such that combining both hemispheres results in 758 parcels (rather than 718).
  • subcortex_atlas_GSR_L.nii - Left hemisphere subcortex-only network partition
  • subcortex_atlas_GSR_R.nii - Right hemisphere subcortex-only network partition
  • subcortex_atlas_GSR_parcels_L.nii - Left hemisphere subcortex-only network parcellation (each parcel separated)
  • subcortex_atlas_GSR_parcels_R.nii - Right hemisphere subcortex-only network parcellation (each parcel separated)
• images directory
  • Figure1.jpg - An illustration of the network partition.
  • cortex_Illustration_ColeAnticevicNetpartition_splithalfvalidation_v1.pdf - An illustration of the split half validation results.
  • ColeAnticevicHumanNetPartition_animation.gif - An animated GIF illustrating the partition across transaxial slices of the S1200 HCP average T1 image.
• NoGSRSubcortex directory
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR.dlabel.nii - Subcortical parcellation without global signal regression (GSR) applied to subcortical voxels as a preprocessing step. Cortex is also included.
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_netassignments_LR.dlabel.nii - Same as above, but at the network assignment level (rather than level of individual parcels).
  • subcortex_atlas_ConjunctionGSRnoGSR_n.dlabel.nii - A version of the parcellation based on the conjunction of the GSR and non-GSR versions, for those who wish to only use subcortical voxels with assignments consistent with and without GSR.
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_netassignments_LR.dscalar.nii - Dscalar version of network assignments.
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR.dscalar.nii - Dscalar version of parcel-level network assignments.
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_LabelKey.txt - Legend text file showing all mappings for the woSubcorGSR parcellation between: Index; KeyValue; Label; Red; Green; Blue; Alpha; Hemisphere; Network; NetworkKey; NetworkSortedOrder, GlasserLabelName
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii - The parcels (same as in CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR.dlabel.nii) sorted by network. The order of parcels within a network is sorted as such: Cortex then Subcortex; L Hemisphere then R Hemisphere then LR. See CortexSubcortex_ColeAnticevic_NetPartition-ReorderedbyNetworks.sh.
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_ReorderedByNetworks.dscalar.nii - The dscalar file corresponding to CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii .
  • CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_v1.1_LR_ReorderedByNetworks.txt - - The labels for CortexSubcortex_ColeAnticevic_NetPartition_woSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii, along with color information (RGBA value).
  • SeparateHemispheres directory - Files with hemispheres separated, for subcortical parcellation without global signal regression (GSR) applied to subcortical voxels as a preprocessing step.
• data directory
  • cortex_fc_avg.pconn.nii - Correlation matrix used for creating the cortical partition. Formatted for visualization in Workbench.
  • meanFCMatSorted.csv - Same correlation matrix, formatted as a comma separated value file.
  • meanFCMatSorted.mat - Same correlation matrix, formatted as a MATLAB file.
  • cortex_gamma1.295_partition_before_reassignment.mat - Network assignment for the cortical parcels prior to the clean up (reassignment) step.
  • CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_netassignments_LR.dscalar.nii - Dscalar version of network assignments.
  • CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dscalar.nii - Dscalar version of parcel-level network assignments.

Network Order Information and Examples for Re-ordering CIFTI Outputs

Note that the file CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii is ordered by cortical then subcortical parcels. Also, note that the subcortex is ordered in L/R order instead of by network (refer to *_LabelKey.txt for label mappings). The key reason for this choice was to provide a clear distinction between cortical and subcortical parcels. Here is an example:

For most people this result may not be the preferred choice, especially if you wish to publish an ordered matrix by network:

Therefore, if you would like to run an analysis such that your order is ordered by network, then the following file should be used on the front end of your analysis:

CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_ReorderedByNetworks.dlabel.nii

If you already ran an analysis that was parcellated with CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii and wish to reorder your result by network then follow the following steps for a ptseries or pscalar file types:

wb_command -cifti-reorder <yourfile>.ptseries.nii COLUMN ColeAnticevicNetPartition/cortex_subcortex_community_order.txt <yourfile>.ptseries.nii

---

If you already ran an analysis that was parcellated with CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii and wish to reorder your result by network then follow the following steps for a pconn file type:

• Note -- you will have to re-order the file twice: once by COLUMN and once by ROW

wb_command -cifti-reorder <yourfile>.pconn.nii COLUMN ColeAnticevicNetPartition/cortex_subcortex_community_order.txt <yourfile>.pconn.nii

wb_command -cifti-reorder <yourfile>.pconn.nii ROW ColeAnticevicNetPartition/cortex_subcortex_community_order.txt <yourfile>.pconn.nii

Volume version of the Glasser surface parcels

We have been receiving many inquiries into using our network partition with volume (rather than CIFTI) data. We defer to the creators of the Glasser et al. (2016) cortical parcellation (https://balsa.wustl.edu/study/show/RVVG) in saying that using a volume version is not recommended. See here for more info on this recommendation: https://www.mail-archive.com/[email protected]/msg03398.html

That said, there are several options for those willing to take the risk in using a volume version of the Glasser cortical parcellation. As recommended previously (https://www.mail-archive.com/[email protected]/msg03398.html) the best option is to use Connectome Workbench to map the group-level cortical surface label file to each subject's individual volume:

1. Use wb_command -cifti-separate to make GIFTI label files
2. Use wb_command -label-to-volume-mapping to map GIFTI label files to the volume

Less ideal but still possible, there are also several options for versions of the Glasser cortical parcellation mapped to group-level volume space. Use with caution. Examples:

• https://cjneurolab.org/2016/11/22/hcp-mmp1-0-volumetric-nifti-masks-in-native-structural-space/
• https://neurovault.org/collections/1549/
• https://figshare.com/articles/HCP-MMP1_0_projected_on_MNI2009a_GM_volumetric_in_NIfTI_format/3501911
• Apparently AFNI now comes with a volume version of the Glasser cortical parcellation, as described here: https://openwetware.org/wiki/Beauchamp:CorticalSurfaceHCP

coleanticevicnetpartition's Issues

Glasser to Cole Anticevic subcortical parcels

How can we map the Glasser parcels to the Cole Anticevic atlas for the subcortical parcels? In CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt, I see that "NETWORKSORTEDORDER" corresponds to the Glasser parcels and "NETWORK" is the mapping to the Cole Anticevic networks, but after row #360 which should start with the subcortical parcels, it is unclear how what network each subcortical parcel corresponds to. It seems as though there are multiple network assignments for the L/R subcortical parcels?

For example, there is "Visual1-01_L-Accumbens"..and so forth for all of the subcortical parcels and then switches to "Somatomotor-01_R-Accumbens", etc.

To pull out individual regions/ networks as masks

Hi,

I would like to save few parcels/ networks as masks to be used to extract fMRI time series from? How can I do it?

Thanks,

Shilpi Modi

Possible to get the labelled cortical-subcortical MNI coordinates corresponding to the 12 networks?

Hello @jielisaji! Thank you so much for this data! :)
I am trying to get the [x,y,z] MNI coordinates corresponding to each of the 12 networks (so something like x,y,z, -> network 1) so I can plot it in a python diagram. May I ask if this dataset is available? If not, may I ask which are the nii cortical-subcortical images where the 12 networks are already labelled?

My friend and I have tried going through the files, the README, the humanconnectome site, and also tried running the LoadParcellatedDataInPython_Example script too but we're not too sure how we can go about getting the coordinates..

Thanks a lot for your time! >.<

Discrepancy in subcortical labels for '*_parcels_LR.dlabel' file and 'data/*parcels_LR.dscalar file'

Hi, this a somewhat minor issue but the subcortical region labels are different for the 'CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii' file and the 'data/CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dscalar.nii.file', even though the documentation implies they should be the same. I think the numbering in the dscalar file is more sensible (it goes from 361-718, whereas dlabel goes from 1010-12170), so maybe the dlabel file should be switched to that scheme?

Labels for separate hemispheres

Dear team,

I really enjoy your parcellation- thank you so much for this! As I would like to use it as ROIs for my tractography, I need to translate them to MNI space which is really hard using connectome workbench (that is suited for cortex data and does not recognize that the .surf.gii-files that you included contain the subcortex - there is an error message saying: "specified file and direction does not contain the requested surface structure 'CEREBELLUM'/BRAIN_STEM etc).

I tried to use the .nii.gz-files from the SeparateHemispheres folder, however the amount of regions is not exactly the same as in LR (left hemisphere: 213 regions, right: 180 regions), so I do not know how to label the regions. Could you provide me with the labels for these files? Thanks!

Using LoadParcellatedDataInMatlab_Example.m

Dear Dr. Cole,

I am looking to apply the ColeAnticevic atlas to my data in order to run a parcellated task analysis. I preprocessed my fMRI data using fmriprep and obtain fmri dtseries in fsLR space.

Can I directly use LoadParcellatedDataInMatlab_Example.m on each subject by using the subject's dtseries as the input file? Is there an additional registration step to be used to make the ColeAnticevic atlas individualized?

parcelTSFilename

Hi,

I want to use the ColeAnticevicNetPartition in an fMRI study to study dynamic functional connectivity of various brain networks during task performance and rest.

1. In LoadParcellatedDataInMatlab_Example.m what is

parcelTSFilename='Output_Atlas_CortSubcort.Parcels.LR.ptseries.nii';

Regards,

Shilpi Modi

Any volumetric version?

Hi! I am using your atlas and want to project the surface atlas back to volumetric space for neurosynth decoding. I know the subcortical ROIs are already in volumetric space but unsure what to do with the cortical parcels. Apologize if this is an easy question.
Thanks,
Oliver

Network mismatch in cortex_parcel_network_assignments.mat

In a previous issue, it was mentioned that the order in cortex_parcel_network_assignments.mat corresponded to the parcel order in Glasser parcellation, in which all the left parcels correspond to the first 180 indices and the right parcels correspond to the last 180 indices. Given that left and right parcels should belong to the same one out of 12 networks, there is a mismatch in networks assigned to a given left parcel and its corresponding right parcel in cortex_parcel_network_assignments.mat.

After loading cortex_parcel_network_assignments.mat into Matlab, I ran this command:
mismatch = []; mismatch(:, 1) = netassignments(1:180, 1); mismatch(:, 2) = netassignments(181:360, 1)

See attached mismatch.txt file, in which the first column corresponds to the 180 left parcels' network IDs and the second column corresponds to the 180 right parcels' network IDs .
mismatch.txt

For example, row 11 col. 1 corresponds to L_Premotor Eye Field and row 11 col. 2 corresponds to R_Premotor Eye Field, however, their network ID #s do not match (the left parcel was assigned to 5 (dorsal-attention) , whereas the right parcel was assigned to 4 (Cingulo-opercular)).

These inconsistencies in network IDs are also in rows 25, 28, 58, 62, 74, 81, 82, 93, 104, 128, 162, and 177.

Could you clarify this?

wb_command -cifti-parcellate

I used the fillowing command:

eval(['!wb_command -cifti-parcellate ' inputFile ' ' parcelCIFTIFile ' COLUMN ' parcelTSFilename ' -method MEAN'])

The output obtained is as:

While running:
wb_command -cifti-parcellate input.dtseries.nii CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii COLUMN Output_Atlas_CortSubcort.Parcels.LR.ptseries.nii -method MEAN

ERROR: input cifti files must have the same volume space

Can someone explain the error to me and the way to run the code. My input dtseries was obtained after processing using fmriprep.

Regards,

Shilpi Modi

Using Cole Atlas parcels for extracting BOLD signal

Hi,

• I have pre-processed my fMRI task data with fMRIPrep and obtained the output dtseries in cifti (fsLR) space.

1. I visualised output dtseries in wb_view and it seems to be registered with the CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii

However, is there a way so that I know that my time series data is properly registered with the atlas and the signal extracted using the below code actually gives the values for corresponding parcels of the brain.

• Thereafter, I used the LoadParcellatedDataMAtlab_Example.m to parcellate my dtseries and obtain BOLD signal from the language network parcels.

2. The 120X36 matrix (120 dynamics and 36 language parcels (I have omitted one)) generated shows a high level of coherence between the BOLD signals from all the 36 parcels for the entire time series even for a task data. Is it expected or is an error?

Thanks and regards,

Shilpi Modi

How to get the parcel name of FCMat_sorted?

Hi all,

I have obtain the FCMat_sorted (718*718) by LoadParcellatedDataInPython_Example.py scripts. In the script, the FCMat was ordered by cortex_subcortex_community_order.txt. I notice there is another file CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt that store the parcel informations. I'm not sure the first row in FCMat_sorted (first parcel in FCMat_sorted) corresponds to 1 of column NETWORKSORTEDORDER or 1 of column INDEX in *LabelKey.txt file.

Maybe my description is a little confusing, in other words, the value in cortex_subcortex_community_order.txt correspond to which column in *LabelKey.txt file? The NETWORKSORTEDORDER or INDEX?

Sorry I am a beginner in fMRI, the description may not be clear, if you have any questions, please reply and I will add it in time.

Weihan

Anatomical location of various parcels

1. In the CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt file, each label/ parcel of a particular network is assigned an index/key value. The hemisphere to which a parcel belongs and whether it’s a cortical/ sub-cortical parcel can be identified. However, is there a way to know exactly to which lobe of the cortex/ anatomical location, a particular parcel belongs.

2. In the supplimentary table 1 of Glasser, 2016 a table with the description of Glasser cortical parcels is provided. Is the index/ key value same for Cole Parcels as well?

Thanks.

Shilpi Modi

Viewing only selected parcels

Hi,

I downloaded the atlas and am able to view various networks/ parcels using the workbench.

Is there a way to visualize only the parcels that we want and not the entire network/ all the 718 parcels.

Thanks.

Shilpi

Missing hemisphere column values for certain subjects in CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt

Using the first column as indices, line 521 shows Dorsal-Attention-01_L_Amgydala. That line and the ones below do not have a column value for hemisphere (though of course the hemisphere can be inferred from the fourth column).

wb_command cifti-parcellate throws error: data file is missing voxel (49, 66, 28), which is used by label 'Orbito-Affective'

Hello,

I am attempting to parcellate MyelinMap.32k_fs_LR.dscalar.nii files based on your ColeAnticevic atlas using the following code: ${CARET7DIR}/wb_command -cifti-parcellate subjid.MyelinMap.32k_fs_LR.dscalar.nii /path_to_my_HCP_templates/CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_netassignments_LR.dlabel.nii COLUMN subjid.testing.pscalar.nii -method MEAN

I get the following error: "data file is missing voxel (49, 66, 28), which is used by label 'Orbito-Affective'"

Could this be related to only inputting a surface file and therefore not having any "voxels"? Is there a purely cortical dscalar file that I might have missed? Also, could you please point me to the corresponding .border file in order to get the correct labels for each region (i.e. using wb_command -border-export-color-table or the like).
Any help would be greatly appreciated!
Kind regards Linn

Cortex only files (L/R hemispheres separate) appear to have flipped numbering

Hi @jielisaji ,

I noticed that a new set of separate cortical hemispheres were recently uploaded (maybe last 8 months or so):
ColeAnticevicNetPartition/SeparateHemispheres/Q1-Q6_RelatedValidation210.R.CorticalAreas_dil_Final_Final_Areas_Group_Colors.32k_fs_LR.dlabel.nii
and
ColeAnticevicNetPartition/SeparateHemispheres/Q1-Q6_RelatedValidation210.L.CorticalAreas_dil_Final_Final_Areas_Group_Colors.32k_fs_LR.dlabel.nii

I know there have been issues with L/R ordering in the past, which is why we have opted to deal with separate cortical hemispheres to preserve L -> R ordering from 1 -> 360. However, I noticed when opening each file separately, e.g.,

rh = nib.load('SeparateHemispheres/Q1-Q6_RelatedValidation210.R.CorticalAreas_dil_Final_Final_Areas_Group_Colors.32k_fs_LR.dlabel.nii').get_data()
lh = nib.load('SeparateHemispheres/Q1-Q6_RelatedValidation210.L.CorticalAreas_dil_Final_Final_Areas_Group_Colors.32k_fs_LR.dlabel.nii').get_data()

that
In [37]: print(np.unique(rh)) [ 0. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180.]

and
In [36]: print(np.unique(lh)) [ 0. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. 288. 289. 290. 291. 292. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310. 311. 312. 313. 314. 315. 316. 317. 318. 319. 320. 321. 322. 323. 324. 325. 326. 327. 328. 329. 330. 331. 332. 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343. 344. 345. 346. 347. 348. 349. 350. 351. 352. 353. 354. 355. 356. 357. 358. 359. 360.]

Which seems the opposite of what I would expect. In the label txt file ('CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR_LabelKey.txt'), it shows LH regions should be numbered 1-180 and RH regions are 181-360. I wonder if this is due to the recently added files (e.g., the 'RelatedValidation' files)? (Or perhaps this was always the case?)

Thanks,
Taku

Mismatch between Glasser Parcels and Cole-Anticevic Communities

There seems to be a mismatch between the order of the Glasser parcels (all left hemisphere regions first - i.e. the first 180 correspond to left regions, the last 180 correspond to right regions) and those in cortex_community_order.mat and cortex_parcel_network_assignments.mat.
GlasserToColeAnticevicMapping.txt

See attached GlasserToColeAnticevicMapping.txt file, in which the Glasser parcels have been reordered in ascending order. For example, parcel #1 and #181 should correspond to the left primary visual cortex and right primary visual cortex, respectively, however, in the mat files you have provided, it shows that parcel #1 belongs to the Cole-Anticevic Network #1, while parcel #181 belongs to Cole-Anticevic Network #6. I believe the right and left hemisphere parcels should belong to the same network, correct?
Glasser2016-supplement-Neuroanatomical_Supplementary_Results.pdf

Starting on page 81 of the Neuroanatomical Supplementary Results from Glasser et. al. 2016, the order of the Glasser parcels are listed in the table. This is the order I am going off of when I refer to a mismatch between the mat files on Github and the Glasser parcels.

Could you provide some clarification on this?

Seeking your advice of template

Dear ColeLab team

This is Woo Sung Kim, a Ph.D. student at Jeonbuk National University from Korea.
I have read your article "Mapping the human brain's cortical-subcortical functional network organization."
After reading your paper, I do have questions about the functional network parcellation mask.
I would be grateful if you can find time for me.

• I want to discuss applying the cortical part of the network partition template used in the paper to my analysis.
  The tool that I currently use for functional connectivity analysis is CONN (https://web.conn-toolbox.org/), so I would like to apply the network partition template in CONN toolbox.
  'Data, software, and the network partition are available here: https:// github.com/ColeLab/ColeAnticevicNetPartition and https://doi.org/10. 5281/zenodo.1455791.'
  When I download the code and files through the address above and checked them, I couldn't find the network partition template.
  I want to discuss whether I should make and use the network partition template using the 'LoadParcellatedDataInMatlab_Example_cortexonly.m' code.

• I understand subcortex template also assigned with 12 cortical networks. Is there any specific detail of the subcortex template, especially the thalamus? Like thalamus is also parcellated to 12 parts or the whole thalamus is included in some of the 12 cortical networks.

• If you can provide the network partition template of the cortical part and subcortex (especially the thalamus) in the atlas (3D brain image; ~.nii.gz / ~.nii) format, then would you please provide it for me?

Thanks for your time and expertise in advance.

Best Wishes

Woo

Parcel naming

Thanks for making this available. It's great to have a rigorous network partition, including the addition of subcortex, of the Glasser parcellation.

However, it's a bit cumbersome to work with the current parcel naming scheme. Specifically, there is one parcel in each network that does not have a numeric suffix as part of its name. e.g., Most parcels in the visual network are named "Visual_#", but "Visual" is the name of one specific parcel as well . So not only does the name of one parcel for each network not map to a standard naming format, but the name of that parcel is identical to the name of the network itself.

Also, it is bit non-intuitive that the cortical parcels are named with an underscore, and the subcortical ones with a dash. A more intuitive distinction in the naming between the cortical and subcortical parcels would be helpful.

In terms of sorting, it would be convenient if the numeric portion of the parcel names was padded to a consistent number of digits.

Last, I know all the necessary information is already contained as part of the *.txt and *.dlabel.nii files, but it is a bit hard to integrate. It would be convenient if a single summary csv was provided that linked the parcel names, the name of the parcel in the original Glasser version, their hemisphere, their numeric position in the dlabel file, the parcels 'key' values, their network assignment, the networks 'key' value, and the parcels position when they are resorted according to the *order.txt files. Without this, it is challenging to navigate between the different data representations, especially since hemisphere is not encoded into the parcel names, and the numeric position in the dlabel file and the 'key' value in the dlabel file are not necessarily the same. Having such a single summary csv would make it much easier to interpret the structure of FC matrices that have been sorted according to the provided *order.txt files.

Mapping from Glasser to Cole Anticevic

In the cortex_parcel_network_assignments.txt file, do the 360 rows correspond to the the exact same order in Glasser parcellation? Or is there a separate file or way to check if the parcel labels match up? I am trying to do a mapping between Glasser parcels and CAB to get the corresponding network labels.

Thanks!

about FC Matrix obtained

Thanks for making this available, great work!

Is there the FC Matrixs with 3D coordinates of MNI space available?
Maybe just like:

x  y  z node_label node_size
-39.5807 -5.71401 51.108	1	1.000000	-
40.2365	-8.38623	52.3802	1	1.000000	-
-19.2128	34.7426	42.474	1	1.000000	-
20.9405	30.7998	43.9931	1	1.000000	-
-17.8767	47.25	-12.8836	1	1.000000	-
17.7395	47.6141	-13.6109	1	1.000000	-
-34.268	32.3246	35.7804	1	1.000000	-
36.5444	32.998	34.1722	1	1.000000	-
-31.9333	50.4556	-9.33333	1	1.000000	-
32.7245	52.3776	-10.5816	1	1.000000	-
-49.3067	13.1779	20.2178	1	1.000000	-
49.3539	14.8219	21.4632	1	1.000000

Each row represents a node information. For example, the (x, y, z) is the 3D coordinate of MNI space.

using GSR atlas on functional data without GSR

Hi,

Firs of all, congratulations for such a wonderful atlas.

One question, is it OK to use the GSR version of your atlas, even if the functional data to which I will apply the atlas was not preprocessed including that step?

Thanks a lot in advance!

Ordering of network labels

In the network_labelfile.txt file, it lists the 12 Cole-Anticevic Networks, but to clarify, does the list order correspond to the network number in increasing order that can be matched up using cortex_parcel_network_assignments.mat?

By this, I mean:

Visual1 - network 1
Visual2 - network 2
Somatomotor - network 3
Cingulo-Opercular - network 4
Language - network 5
Default - network 6
Frontoparietal - network 7
Auditory - network 8
Posterior-Multimodal - network 9
Dorsal-attention - network 10
Ventral-Multimodal - network 11
Orbito-Affective -network 12

Not valid .nii files?

Thanks for the parcellations, great work!

What files exactly are these?

CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_netassignments_LR.dlabel.nii 
CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii

Despite the extension, they do not appears to be valid NIfTi's.

Error message by load_nii.m:
"Error using load_nii_hdr (line 77)
File "CortexSubcortex_ColeAnticevic_NetPartition_wSubcorGSR_parcels_LR.dlabel.nii"
is corrupted."

Error message by MRIcron:
"Warning: the header file is not in NIfTi format [the first 4 bytes do not have the value 348]. Assuming big-endian data."