Heterogeneous, temporally consistent, and plastic brain development after preterm birth

This repository accompanies the paper:

Melissa Thalhammer, Jakob Seidlitz, Antonia Neubauer, Aurore Menegaux, Benita Schmitz-Koep, Maria A. Di Biase, Julia Schulz, Lena Dorfschmidt, Richard A. I. Bethlehem, Aaron Alexander-Bloch, Chris Adamson, Gareth Ball, Claus Zimmer, Marcel Daamen, Henning Boecker, Peter Bartmann, Dieter Wolke, Dennis M. Hedderich, and Christian Sorg (2024). Heterogeneous, temporally consistent, and plastic brain development after preterm birth.

Abstract

The current view of neurodevelopment after preterm birth presents a strong paradox: diverse neurocognitive outcomes suggest heterogeneous neurodevelopment, yet numerous brain imaging studies focusing on average dysmaturation imply largely uniform aberrations across individuals. Here we show both, spatially heterogeneous individual brain abnormality patterns (IBAPs) but with consistent underlying biological mechanisms of injury and plasticity. Using cross-sectional structural magnetic resonance imaging data from preterm neonates and longitudinal data from preterm children and adults in a normative reference framework, we demonstrate that brain development after preterm birth is highly heterogeneous in both severity and patterns of deviations. Individual brain abnormalities were also consistent for their extent and location along the life course, associated with glial cell underpinnings, and plastic for influences of the early social environment. Thus, IBAPs of preterm birth are spatially heterogenous, temporally consistent for extent, spatial location, and cellular underpinnings, and plastic for social-environmental impacts. Our findings extend conventional views of preterm neurodevelopment, revealing a nuanced landscape of individual variation, with consistent commonalities between subjects. This integrated perspective of preterm neurodevelopment implies more targeted theranostic intervention strategies, specifically integrating brain charts and imaging at birth, as well as social interventions during early development.

License and citation

Content of this repository is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Please cite the publication when using the data or code provided in this repository. Third-party data included in this repository requires direct citation if used in other contexts. This applies to all MRI imaging data, normative models, or gene expression data. See Methods and Supplementary Information of the publication for details and resources.

Installation

This repository contains the code to reproduce the analyses and figures of the publication. The code is written mainly in Python and R, only Fig. 5 is generated in MATLAB. The installation steps are described below. Analyses were carried out on a MacBook Air with macOS Sonoma 14.6.1 and might require adjustments for other operating systems, especially Windows. The code was additionally tested on a HP Pro notebook with Windows 11 Pro. The installation command differ slightly for Windows, please individually adjust the commands accordingly.

Software requirements

• Python 3.11.9
• R 4.4.1
• MATLAB 2023b (for Fig. 5 only)

Installation steps

1. Create an empty directory and navigate into it:

   mkdir /path/to/your/setup && cd /path/to/your/setup

2. Clone this repository:

   git clone https://github.com/Melissa1909/preterm-brain-heterogeneity 

3. Clone the BrainChart repository, copy it into /code, and remove unnecessary models of original publications:

   cd preterm-brain-heterogeneity/code
   git clone https://github.com/brainchart/Lifespan BrainCharts
   rm -rf BrainCharts/Share
   cd ..

4. Create a virtual environment with Python 3.11.9:

   python3.11 -m venv .venv 

5. Activate the virtual environment:

   On Windows: bash .venv\Scripts\activate

   On macOS/Linux: bash source .venv/bin/activate

6. Install python dependencies:

   pip install -r requirements.txt

7. Install the required R packages (run this in an R environment):

   install.packages(c("ggplot2", "gamlss", "tidyverse", "devtools", "invgamma"))
   devtools::install_github("jcbeer/longCombat")

   Furthermore, install the PROCESS toolbox for R. You can download it here, unzip it, and place it into /code as well.

8. Install the required MATLAB package (for Fig. 5 only):

   cd preterm-brain-heterogeneity/code
   git clone https://github.com/dutchconnectomelab/Simple-Brain-Plot 

Usage

/data

The subfolders would contain the data used in the publication but are empty due to data protection regulations.

• dHCP: Neonatal imaging data from the developmental Human Connectome Project. The dataset requires access application here. Furthermore, data need to be parcellated with the M-CRIB-S parcellation, which corresponds to the adult Desikan-Killiany atlas.
• ABCD: Children imaging data from the Adolescence Brain Cognitive Development Study. Apply to access here and download the data from release 4.0.
• BLS: Adult imaging, behavioral, and perinatal data collected in-house. Access can be granted only upon reasonable request from the authors.
• cell_types: Copy of compiled cell-specific gene set list from all available large-scale single-cell studies of the adult human cortex obtained from the raw Seidlitz et al. dataset.

Jupyter notebooks

The Jupyter notebooks contain all the analyses code in sequential order to reproduce the results of the publication. The notebooks are organized as follows:

1. 0.1_dataPrep_*.ipynb: Preparation of meta- and imaging data for the respective dataset, including quality control, averaging across hemispheres, data harmonization (if applicable), preparation of the input file for subsequent normative modeling.
2. 1_spatialHeterogeneity.ipynb: Contrasting average dysmaturation outcomes (based on group comparisons with two-sample t-tests) and individual brain abnormality patterns (IBAPs) obtained via normative modeling. We show that brain aberrations after preterm birth are spatially heterogeneous between individuals in all three datasets. This script should be run for all datasets and the modalities cortical thickness CT and surface area SA.
3. 2_macrosaleConsistency.ipynb: Investigating the temporal consistency of individual brain abnormality patterns (IBAPs) across the life course. We demonstrate that the extent and location of brain abnormalities are consistent in individual development.
4. 3_microscaleConsistency.ipynb: Investigating the cellular underpinnings of IBAPs and their consistency across the life course. The analysis shows that glial cells underpin adult IBAPs, which is complementary to postmortem findings of glial damage in preterm neonates.
5. 4_developmentalPlasticity_IQ.ipynb: Assessing the plasticity of IBAPs in response to the early social environment and effects of adult IBAPs on cognitive outcome variability. We show that the early social environment influences the severity of brain abnormalities in preterm individuals and that the severity of IBAPs is linked to IQ in adulthood.
6. Supp_Rutherford.ipynb: Reproducing the spatial heterogeneity results in the BLS dataset using a different pre-trained normative model by Rutherford et al. (2022).

MATLAB script

Parts of Figure 5 were generated with Fig5_concept.m in MATLAB 2023b using Simple-Brain-Plot.

/code

Contains additional functions for analyses, which are imported in the Jupyter notebooks.

/outputs

Contains the results of the analyses, including figures.

Troubleshooting

Installation issues

• If you encounter issues with installing the ENIGMA toolbox, please try to install the toolbox again manually as described in the ENIGMA documentation. Important: vtk==9.2.6 is required, I've encountered problems with other versions of vtk.
• Make sure Python and R installations are set in the PATH environment variable. If you encounter issues with the R packages, try to install them manually in RStudio or R console.

Support and contact

If you run into issues, please open an issue on this repository or contact me, including your operating system and a detailed description of the problem. Feedback is appreciated!