Fear_Conditioning_MegaAnalysis_NormModelling

This repository contains code for the (forthcoming) manuscript titled: Neural Correlates of Human Fear Conditioning and Sources of Variability: A Mega-Analysis and Normative Modeling Study of fMRI Data from 2,199 Individuals

CODE:

STAI harmonization:

Harmonization of STAI scores across countries using combat.enigma version 1.1.1 (https://CRAN.R-project.org/package=combat.enigma):

• r_codes/stai_harmonization.R

Mega-Analysis Code:

Mega-analysis using combat.enigma version 1.1.1 (https://CRAN.R-project.org/package=combat.enigma). Note that most of the code developed for this project was included in the combat.enigma R package as simple functions for other groups to use, and r_codes/mega_analysis.R only includes the calls to these functions. The contents of the package are fully available at https://cran.r-project.org/src/contrib/combat.enigma_1.1.1.tar.gz

• r_codes/mega_analysis.R

Normative Modelling Code:

Normative Models: Building models + test and evaluate on control test sample

• 01_prepare_wholebrain_model_wmvol_control_test.ipynb
• 02_run_wholebrain_model_wmvol_control_test.ipynb
• 03_evaluate_wholebrain_model_wmvol_control_test.ipynb
• 04_NPM_test_controls.ipynb

Normative Models: Reference cohort - structure coefficients:

• 05_structure_coefficients_control_test.py
• 06_regression_coefficient_weights_control_test.py

Normative Models: Clinical test sample (controls + patients) - test and evaluate:

• 07_prepare_wholebrain_model_wmvol_clinical_test.ipynb
• 08_predict_wholebrain_model_wmvol_clinical_test.ipynb
• 09_evaluate_wholebrain_model_wmvol_clincial_test.ipynb
• 10_NPM_test_controls.ipynb
• 10_split_Z_est_diagnosis_NPM_threshold_combine.ipynb

Compare groups:

• 11_frequency_per_group_compare.ipynb
• 12_frequency_per_diagnosis_compare.ipynb
• 13_barplots_allgroups.ipynb
• 14_Patient_status.ipynb
• 14_Patient_status_permutations.py
• 14_Patient_status_significance.py
• 15_Primary_Diagnosis.ipynb

Validation:

• 16_Contingency_Awareness.ipynb
• 16_Contingency_Awareness_permutations.py
• 16_Contingency_Awareness_significance.py
• 17_Anxiety_scales_elastic_net.ipynb
• 17_Anxiety_scales_elastic_net_significance.py
• 18_Depression_scales_elastic_net.ipynb
• 18_Depression_scales_elastic_net_permutations.py
• 18_Depression_scales_elastic_net_significance.py

NUMERICAL SOURCE DATA:

Mega-Analysis Fies:

Normative Modelling Files:

The following outputs were generated for the control_test and clinical_test:

• EV_clinical_test.nii.gz
• EV_test.nii.gz
• kurtosis_clinical_test.nii.gz
• kurtosis_test.nii.gz
• skew_clinical_test.nii.gz
• skew_test.nii.gz
• SMSE_clinical_test.nii.gz
• SMSE_test.nii.gz

Normative Probability Maps (i.e number of participants with large deviations):

• Freq_neg_dev_controls.nii.gz
• Freq_neg_dev_gad.nii.gz
• Freq_neg_dev_ocd.nii.gz
• Freq_neg_dev_patients.nii.gz
• Freq_neg_dev_ptsd.nii.gz
• Freq_neg_dev_sadnii.gz
• Freq_pos_dev_controls.nii.gz
• Freq_pos_dev_gad.nii.gz
• Freq_pos_dev_ocd.nii.gz
• Freq_pos_dev_patients.nii.gz
• Freq_pos_dev_ptsd.nii.gz
• Freq_pos_dev_sad.nii.gz

Influence of task variables:

• The structure coefficient maps (Correlation coefficients (rho) thresholded by the respective coefficients of determination (rho2>0.3)) for each input variable are included.
• The regression coefficient weights (w/wt/wtn) for each input variable are included.

The data for the Normative Probability Map Count comparisons (Figure 6) are in:

• counts_hist_affective_only.csv
• counts_per_diagnosis_affective_only.csv