This repository contains the analysis code to reproduce the analyses presented in the paper Bathelt, J., Johnson, A., Zhang, M., the CALM team & Astle, D. E. "Data-driven brain types and their cognitive consequences". The main analyses can be found in the following IPython Notebooks:

• Brain_Types_main_analysis.ipynb:

  • descriptive statistics of all samples
  • grouping based on FA values in 20 white matter tracts using community detection
  • application of these grouping to the CALM and ACE sample
  • robustness testing of community clustering with simulated data
  • statistical comparison of cognitive scores between the clustering-defined groups

• BrainTypes_CingulumConnectivity.ipynb:

  • mapping of structural connectivity of the cingulum
  • comparison of cingulum connectivity between the brain types
  • regression analysis of single cingulum connectivity and cognitive scores

• BrainTypes_rsfMRI_analysis.ipynb:

  • identification of the default mode network (DMN) from group-level independent component analysis (ICA)
  • identification of the DMN from individual ICA
  • comparison of spatial extent and total acitvation of the DMN between brain types

There are also several python scripts that contain the pipelines and functions used for data preprocessing. Specifically:

• BrainTypes_additional_interfaces.py: definition of additional interfaces called by the pipelines

• BrainTypes_additional_pipelines.py: definition of additional pipelines called by higher-order pipelines

• BrainTypes_analysis_functions.py: helper functions used in the analysis notebooks

• BrainTypes_dwi_preproc.py: pipeline for preprocessing of diffusion-weighted data

• BrainTypes_dwi_preproc_submission.py: wrapper script to submit the dwi preprocessing to a compute cluster

• BrainTypes_mean_connectivity_MNI_space.py: functions to map individual cingulum connectivity to MNI space for visualization

• BrainTypes_nki_quality_control.py: script to extract dwi quality indices in the NKI sample

• BrainTypes_rsfMRI_correlation.py: script to extract partial correlations between ROIs from fMRI data

• BrainTypes_rsfMRI_data_quality.py: script to extract quality metrics from fMRI data

• BrainTypes_rsfMRI_IC_matching.py: script to match components from individual ICA to a template

• BrainTypes_rsfMRI_IC_matching_submission.py: wrapper script to submit IC matching to a compute cluster

• BrainTypes_rsfMRI_processing.py: preprocessing pipeline for fMRI data

• BrainTypes_rsfMRI_processing_submission.py: wrapper script to submit rsfMRI preprocessing to a compute cluster

• BrainTypes_Tract_Extraction.py: script to extract FA values defined in an atlas parcellation from dwi

• BrainTypes_whole_brain_tractography.py: pipeline to perform whole-brain probabilistic tractography

• BrainTypes_whole_brain_tractography_submission.py: wrapper script to submit whole-brain tractography to a compute cluster

written by Dr Joe Bathelt MRC Cognition & Brain Sciences Unit University of Cambridge

please direct any queries to: joe.bathelt [at] mrc-cbu.cam.ac.uk