Whole-cortex in situ sequencing reveals peripheral input-dependent cellular and area identity - pilot brain
Description
BARseq was used as a standalone spatial transcriptomic method to interrogate the expression of 104 cell type maker genes in one hemisphere of the mouse cortex. Data consist of 40 hemi-coronal sections covering the whole cortex and subcortical areas. Include intermediate processed data and cell type spatial analysis. Associated clustering and gene expression-level analyses are provided at https://github.com/gillislab/barseq_analysis. Accompanying manuscript: https://doi.org/10.1101/2022.11.06.515380
Files
Steps to reproduce
"Initial_analysis" include scripts for removing overlapping cells, cluster assignment, CCF registration (requires QuickNII and Visualign), and basic QC (excitatory vs inhibitory neuron markers, marker gene expressions, batch effect, cell type laminar distribution). "CCF_registration" contains output of CCF registration using QuickNII and Visualign and a python script that transforms cells into CCF coordinates. "spatial_analysis_2017CCF" contains scripts for cubelet-level cell type analysis. MATLAB_functions: custom matlab functions needed by all analysis code. System requirements: The analysis code was run on a Windows 10 machine with MATLAB 2021b with image processing toolbox, parallel computing toolbox, statistics and machine learning toolbox, and bioinformatics toolbox. The image processing code was run on a Windows 10 machine with MATLAB 2021b with the same toolboxes as described above, Cellpose 2.2, bardensr 0.3.1, n2v 0.3.1, QuickNII 2.2 for Windows, Visualign 0.9 for Windows. Installation guide: Please follow installation instructions for MATLAB, Cellpose, bardensr, n2v, QuickNII, and Visualign. After installing, please unpack the matlab functions MATLAB_functions.zip and add the unpacked files to matlab path. Typical install time should not take more than 2 hours. Instructions: For all scripts, run scripts sequentially following alphanumeric order. Initial_analysis requires data files in "processed_data" To further explore the data, it is easiest to start from "wholecortexBARseq_filt_neurons_fixedbent_CCF.mat", which include both the cell-level data after CCF registration and clustering and cubelet-level data. "variables.docx" provides a description of the data in this file. processed_data/XC119Ldata-20210630-freeze-clustid.mat also contains rolony-level data.