Code supporting Ball et al (2020) biorXiv Cortical morphology at birth reflects spatio-temporal patterns of gene expression in the fetal brain

Python 3.7.3
Required packages include: numpy, scipy, scikit-learn, statsmodels, seaborn

All installed packages are shown in req.txt To clone environment try: conda create -n new environment --file req.txt

R 3.6.2
Required libraries: mgcv, tidyverse, WCGNA, RColorBrewer, vegan

Neuroimaging
Neuroimaging data can be accessed via the Developing Human Connectome Project

RNA-Seq
Processed transcriptomic data can be accessed from development.psychencode.org

GEO accession for other datasets:
BrainCloud GSE30272
Mouse data GSE89998
Single-cell RNA GSE103723

Preprocessed imaging, transcriptomic and validation data can be found in data/.

Python scripts should be run in order, i.e.:
python A__principal_components.py
python B__run_gene_models.py etc

Apart from some larger files that will need to be regenerated, the output of most scripts are already in results/

Jupyter notebooks to reproduce main and supplemental figures are in figures/ with the exception of network visualisations. For these, one can load the appropriate *.graphml files in Cytoscape.

For further details on preprocessing and analysis, please see:

G. Ball, J. Seidlitz, J. O’Muircheartaigh, R. Dimitrova, D. Fenchel, A. Makropoulos, D. Christiaens, A. Schuh, J. Passerat-Palmbach, J. Hutter, L. Cordero-Grande, E. Hughes, A. Price, J.V. Hajnal, D. Rueckert, E.C. Robinson, A.D. Edwards Cortical morphology at birth reflects spatio-temporal patterns of gene expression in the fetal human brain. biorxiv. 2020.