Code repository for the paper, "Tractometry of the Human Connectome Project: Resources and Insights"

Prediction Analysis

Environment setup:

1. Install R packages used in these analyses from prediction_analysis/renv.lock using renv::restore().

2. Change the paths in prediction_analysis/config.R to point towards the data.

3. The scripts prediction_analysis/tract_models.R and prediction_analysis/local_connectome_models.R run predictive models on tract profiles generated by pyAFQ and local connectomes generated by DSI Studio respectively.

4. These predictive models can be found in prediction_analysis/pRedict/R/run_model.R. Details about model specifics can be found in Kruper et al., 2024.

Data Download

Phenotypic data can be acquired from HCP. Most of the predictors used in this analysis are part of the Open Access release, but Age_in_Yrs is considered restricted and requires additional steps.

Local connectome data can be acquired from the data shared as part of Rasero et al., 2021, here: https://figshare.com/s/b97d2d1ba359e6458cb5

Tractometry data can be downloaded from Open Neuro Data, as detailed in Kruper et al., 2024.

Reproducibility details​​

The derivatives were produced using pyAFQ version 0.7.2.dev11163053115, which corresponds to this commit ​ And Azure Batch pyAFQ ​ Support for this project was provided through grant 1RF1MH121868-01 from the National Institutes for Mental Health/The BRAIN Initiative and through an Azure Cloud Credits grant from Microsoft and the University of Washington eScience Institute.