ABCD Resting Neurocognition

Code to support

Sripada, C., Angstadt, M., Rutherford, S., Taxali, A., Clark, D.A., Greathouse, T., Weigard, A., Hyde, L., Heitzeg, M. 2020. Brain Connectivity Patterns in Children Linked to Neurocognitive Abilities. bioRxiv doi: 10.1101/2020.09.10.291500

https://www.biorxiv.org/content/10.1101/2020.09.10.291500v1

Relies on functions from https://www.github.com/mangstad/Misc_utils

Data files necessary for this analysis should eventually be available through NDAR, will update when that study is live.

Instructions for setup

1. Clone this repository to your computer. git clone https://github.com/SripadaLab/ABCD_Resting_Neurocognition.git ./ABCD_Resting_Neurocognition

2. Clone https://www.github.com/mangstad/Misc_utils to your computer git clone https://www.github.com/mangstad/Misc_utils ./Misc_utils

3. Edit paths in s01_Load.m and s04_LME.R to point to the correct locations for the scripts and the data.

4. Look at and/or run the example.m script to get an idea of the steps that happen in the different scripts.

General overview

In very general terms, our procedure is the following: within folds of cross validation

1. do PCA in training data
2. use PCA scores in regression model to explain phenotype
3. apply model to test data and calculate performance

outside folds

1. calculate mean performance over folds
2. fit PCA and regression in full data to generate a consensus connectome