ET_biomarker

This repository contains codes for the MNI ET data set analysis. Matching and confounder are the 2 main tools used in this analysis.

(More details will be added latter.)

Pre-registration report:

Pre-registration report on OSF;

Approved Pre-registration report from Sci. Rep. on figshare

Requirements

Dependencies :

• nibabel>=3.1
• numpy>=1.18
• matplotlib
• pandas
• scipy
• scikit-learn
• statsmodels>=0.12.0
• pydicom>=2.1.2

Installation (need update)

First, make sure you have installed all the dependencies listed above. Then you can install cog-align by running the following commands:

git clone https://github.com/neurodatascience/ET_biomarker
cd ET_biomarker
pip install -e .

You can confirm that the package has successfully installed by opening a Python terminal and running the following commands:

import ET_biomarker

Getting started

The main analysis codes are organized as jupyter notebooks located in the root folder of ET_biomarker, and they are:

0. 0_power_analysis.ipynb: The power analysis for this project;
1. 1_cohort_matching.ipynb: The matching procedure to make sure that the ET and control groups are sex and age matched;
2. 2_analysis_cerebellar_roi.ipynb: The freesurfer and SUIT results analysis;
3. TBD...

-1) Execute a file directly in shell, codes are located in experiments folder (which includes code to re-execute all of the main and supplemental experiments included in the manuscript):

python experiments/exp1.py