Normative neurodevelopment

This repository includes code used to analyze the relationship between dimensional psychopathology phenotypes and deviations from normative neurodevelopment in the Philadelphia Neurodevelopmental Cohort.

Environment build

conda create -n normative_neurodev_cs_t1 python=3.7
conda activate normative_neurodev_cs_t1

# Essentials
pip install jupyterlab ipython pandas numpy seaborn matplotlib nibabel glob3 nilearn ipywidgets
pip install jupyter_contrib_nbextensions && jupyter contrib nbextension install

# Statistics
pip install scipy statsmodels sklearn pingouin pygam brainspace bctpy

# Extra
pip install mat73

# Pysurfer for plotting
pip install vtk==8.1.2
pip install mayavi
pip install PyQt5
jupyter nbextension install --py mayavi --user
jupyter nbextension enable --py mayavi --user
jupyter nbextension enable --py widgetsnbextension
pip install pysurfer

cd /Users/lindenmp/Google-Drive-Penn/work/research_projects/normative_neurodev_cs_t1
conda env export > environment.yml
pip freeze > requirements.txt

Code

In the code subdirectory you will find the following Jupyter notebooks and .py scripts:

1. Pre-normative modeling scripts:

• 0_get_sample.ipynb
  • Performs initial ingest of PNC demographic data, participant exclusion based on various quality control.
  • Produces Figures 2A and 2B.
  • Designates train/test split.
• 1_compute_node_metrics.ipynb
  • Reads in neuroimaging data.
  • Sets up feature table of regional brain features.
• 2_prepare_normative.ipynb
  • Prepares input files for normative modeling.
• 3_results_sample_characteristics.ipynb
  • Characterizing sample demographics
  • Produces Figures S1, S2, S3, and S9

2. Run normative modeling:

• cluster/4_run_normative.py
  • Submits normative models to the cbica cluster

3. Results:

• 5_results_forward.ipynb
  • Plots predictions from the normative model as annualized percent change.
  • Produces Figures S4
• 6_results_correlations.ipynb
  • Computes regional correlations between psychopathology dimensions and deviations from the normative model
  • Produces Figures 2, 3, and S8
• 7_results_case_control.ipynb
  • Computes regional Cohen's d comparing deviations from depression and ADHD groups against healthy controls
  • Computes spatial correlation of Cohen's d values between depression and ADHD groups
  • Repeats above analyses controlling for overall psychopathology
  • Produces Figure 4

4. Prediction:

• 8_prepare_prediction.ipynb
  • Prepares input files for prediction modeling.
  • Produces Figures S5, S6, and S7
• cluster/9_job_submitter.ipynb
  • Submits prediction models to the cbica cluster

5. Results (cont.):

• 10_results_model_performance.ipynb
  • Produces Figures 1 and S10