Skip to content

rob-mccutcheon/wang_model

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

WANG MODEL

Python 3 translation of the circuit model and fitting algorithm originally in matlab at: https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/fMRI_dynamics/Wang2018_MFMem

NB functional connectivity matrices should be supplied containing untransformed pearson coeficients, r-to-z is done within the script.

DATASETS

HCP TEST-RETEST

  1. Test and retest data downloaded from https://db.humanconnectome.org/ to NaN $HOME/e_i_modelling/data/hcp_test_retest
  2. Processing code is on the NaN $HOME/e_i_modelling/code/hcp_test_retest: a)unzip.py, b)demean.sh c) parcellate.py
  3. Connnectivity matrices transferred to rosalind form NaN with scp

HCP EP

Data processed using HCP pipeline, move it from the HCPEP/imagingcolelction01 to processed befoer doing scripts scripts are at $HOME/hcp/code/HCPpipelines-4.2.0-rc.1/Examples/Scripts/

I was able to get the gradient coefficients file off the steiner prisma

In terms of modules you generally want freesurfer 6.0.0, fsl 6.0.1 and connectomewb 1.4.2. Matlab depends on the scripts

  1. SetUpHCPPipeline.sh: Amend to link to required code (e.g. need to have downloaded Fix)
  2. Prefreesurfer - need to be in a python environment with gradunwarp installed (r_env has this)
  3. Freesurfer
  4. Post Freesurfer
  5. Generic Volume - may need to be in python environment with gradunwapr (e.g. r_env)
  6. Generic Surface
  7. ICA Fix - use matlab runtime, make sure that anaconda 3.7 is loaded and that matlab unloaded, need to set enough memory: setenv FSLQUEUE_OPTS "-l h_vmem=40G". Fix settings.sh needs to have: FSL_FIX_MCRROOT="${HOME}/hcp/code/matlabruntime/mr". Seems that evenn with this sometimes wrong verison of R loaded - try installing a package perhaps if not working, should be R 4.0
  8. Post Fix - use matlab module - make sure matlab 9.5.0 loaded
  9. MSMAll
  10. Dedrift and resample
  11. Parcellate

Fix

I am running it using matlab compiled runtime (download from mathworks), and r that is in anaconda 3.7, nneed to have installed the specific packages specifiedin the fix readme using devtools install_version

Specific subjects

1007, 1011 - missinng a unprocessed tw1 so excluded 1006, 1001 - spin echo field maps had to be renamed from '2' to '1'

1031_01_MR 1032_01_MR 1033_01_MR 1034_01_MR 1035_01_MR 1036_01_MR 1037_01_MR 1038_01_MR 1039_01_MR 1040_01_MR 1041_01_MR 1043_01_MR 1044_01_MR 1045_01_MR 1047_01_MR 1048_01_MR 1050_01_MR 1051_01_MR 1052_01_MR 1053_01_MR 1054_01_MR 1056_01_MR 1057_01_MR 1060_01_MR 1061_01_MR 1063_01_MR 1064_01_MR 1065_01_MR 1066_01_MR 1067_01_MR 1068_01_MR 1069_01_MR 1070_01_MR 1071_01_MR 1072_01_MR 1073_01_MR 1074_01_MR 1075_01_MR 1076_01_MR 1077_01_MR 1078_01_MR 1079_01_MR 1080_01_MR 1081_01_MR 1082_01_MR 1083_01_MR 1084_01_MR 1085_01_MR 1086_01_MR 1087_01_MR 1088_01_MR 1089_01_MR 1091_01_MR 1093_01_MR 1094_01_MR 1095_01_MR 1098_01_MR 1099_01_MR

About

Python translation of the circuit model and fitting algorithm originally in matlab at: https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/fMRI_dynamics/Wang2018_MFMem

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published