Description

Methods (English)

scRNA-seq analysis

Primary contact: sandmann@dnli.com

Rendered outputs & embedded R code

The scRNAseq_analysis_reports.zip archive contains a website generated via literate programming using the quarto engine. Please download the scRNAseq_analysis_reports.zip file, decompress it and open the index.html file it contains in your web browser.

By browsing to the Analyses tab, you can see the R code and the figures it generates side-by-side for each of the figures and supplementary figures describing the single-cell RNA-seq data.

Reproducing the analysis results

Primary data analysis with cellranger

Raw single-cell RNA-seq data (FASTQ files) is available in the NCBI GEO repository (Series GSE262436). Raw and processed count matrices for each sample were generated with cellranger count (version v7.1.0) available from 10X Genomics. 

{targets} workflow

All downstream processing was performed using R (version 4.4.1). The multi-step analysis was implemented using the targets R package.

The scRNAseq_analysis_r_package.zip file available in this repository contains an R analysis package that includes (among other files):

1. The _targets.R file that orchestrates the analysis workflow via the targets R package.
2. Custom functions used by _targets.R in the R subdirectory
3. A serialized R object used to assign cell type labels: data/reference.rda
4. A serialized R object containing a random forest classifier to distinguish cells from male and female animals: data/sex_predictor.rda
5. Multiple quarto markdown (.qmd) files in the vignettes subdirectory that load the output of the _targets.R workflow and create the final analyses and figures reported in the manuscript.

Each of the quarto markdown files loads one or more precomputed outputs, stored in the _targets folder. This folder can either be generated from scratch (see section "{targets} workflow" below), or you can download our precomputed results in the _targets.zip file in this repository (see section "Precomputed output" below).

Note: The scRNAseq_analysis_r_package analysis R package can also be installed as an R package, but that is not required to reproduce our analysis.

Dependencies 

To reproduce our analysis, you not only need our code but also the same versions of numerous R packages. The scRNAseq_analysis_r_package folder you extracted above contains the renv.lock file that can be used to restore the same analysis environment with the renv R package. (It lists all R packages that were installed on our system, as well as their exact version numbers.)

Please note that we used R version 4.4.1 and refer to the documentation of the renv::restore() function for further details.

Rerunning the secondary analysis

Once you have restored the renv environment on your local system, you can rerun our analysis pipeline from stratch by

1. Starting a fresh R session (with the renv environment you restored above)
2. Setting your working directory to the analysis R package (e.g. the same directory that contains the _targets.R file).
3. Using targets to start the workflow: targets::tar_make()

Please note that some of the analysis steps require a large amount of RAM; we completed the workflow on a virtual machine running Ubuntu 22.04.5 LTS linux with 8 cores and 64 Gb of RAM.

Precomputed output

The targets workflow stores intermediate and final output files in the _targets subfolder. For convenience and to facilitate reproducibility, we have included the  _targets.zip archive in this repository as a separate zip archive.

To use our intermediate & final R objects, please download and decompress the _targets.zip file, and then place the resulting _targets folder into the root of the scRNAseq_analysis_r_package folder with the analysis code (see above), e.g. at the same level as the _targets.R script.

Methods (English)

Image analyses

Primary contact: djoy@dnli.com

FIJI/ImageJ

Macros to reproduce the analyses performed in FIJI/ImageJ are provided in the fiji_macros.zip archive. See the README.md file in that archive for a description of the scripts and how to use them. 

Allen Mouse Brain Atlas

We provide a Zarr image archive format of the Allen Mouse Brain Atlas in AllenBrainAtlas.zarr.zip containing our additional fused regions as described in Khoury et al 2024. The original Nissl stained volumes and annotations were downloaded from https://download.alleninstitute.org/informatics-archive/converted_mouse_ccf/. If you use this atlas in your work, please cite the original manuscript (DOI: 10.1016/j.cell.2020.04.007)

Wang, Q. et al. The Allen Mouse Brain Common Coordinate Framework: A 3D Reference Atlas. Cell 181, 936-953.e20 (2020).

See the README.md file in khoury_2024_registration.zip for instructions and example code using this atlas.

Longitudinal Cynomolgus Macaque Atlas

We provide code to use the Longitudinal Cynomolgus Macaque Atlas to analyze light sheet images of macaque brains. To use this code, you must first download the longitudinal atlas from https://www.nitrc.org/projects/cyno_4d_atlas/. If you use this atlas in your work, please cite the original manuscript (DOI: 10.1016/j.neuroimage.2021.118799)

Zhong, T. et al. Longitudinal brain atlases of early developing cynomolgus macaques from birth to 48 months of age. NeuroImage 247, 118799 (2022).

See the README.md file in khoury_2024_registration.zip for instructions and example code using this atlas.

Brain Registration Jupyter Notebooks

We provide two Jupyter notebooks to register and calculate mean and depthwise profiles of fluorescence intensity in mouse and macaque brains:

• CynoBrainRegistration.ipynb - Register macaque brains to the macaque brain atlas and extract regional statistics
• MouseBrainRegistration.ipynb - Register mouse brains to the mouse brain atlas and extract regional statistics

See the README.md file in khoury_2024_registration.zip for installation instructions and usage.