Skip to content

HAM-lab-Otago-University/Commonality-analysis-ABCD4.0

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Commonality-analysis-ABCD4.0

It is assumed that the readers of the codes have ABCD data tables from the internet. So all of the R codes start by processing the raw tables downloaded from ABCD webpage

  1. process the five cognitive tasks and compute gfactor. The code can be found in computing gfactor.

  2. Modelling the brain scan data. Starts with cleaning the raw data tables with quality control (codes in the folder called cleaning raw table brain scan). Then all the scan features are processed in the script called scan_feature_processing for the elastic net fitting. Fitting the elastic net is done on a supercomputer called Nesi and the codes are in brain_scan_modelling_nesi (gfactor_DTI_modelling is an example of elastic net model fitting). There are some problems with DTI output collecting and those outputs are processed separately

The predictive performance and feature importance of elastic net can be found in individual_brain_modalities_scatterplot and feature_importance_enet_brain_continuous_palettes respectively.

The folder with the name stacking_brain_scan_models has three files. random_forest_stacking_data_prep is the script to prepare the elastic net predictions to fit the stacking random forest model. random_forest_loaded_shap_plots are the feature importance plots of all the sets of brain features. stacking_random_forest_performance is the performance metric of the random forest stacking models.

The model fitting and feature importance computation (Shapley values) are in the Nesi folder as that computation is done by Nesi.

  1. modelling mental health. All the codes are in the folder called Psychopathology

  2. modelling the social demographic lifestyle developmental. All the codes are in the folder called ses_developmental.

  3. modelling the genetics that are related to cognition. All the codes are in the folder called Genetics.

  4. commonality analysis. The commonality analyses are done with linear mixed models common_analysis_cross_sites_models_with_all_features is the analysis with all four features common_analysis_gene_mental_mixed, common_mixed_psy_brain, common_mixed_psy_ses are the commonality analysis between mental health and brain MRI, social demographic lifestyle developmental, and genes. common_each_brain_mental_mixed is the commonality between each set of brain scan features and the mental health variables.

scatterplot_predicted_real is a file with scatterplots of all the brain, mental health, social demographic lifestyle developmental, and genes. Also, this file has a table of metrics and three different sets of mental health features.

All the functions needed for the analysis are in the R_functions.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages