Scripts for analyzing the 10X 1.3 million brain cell data

The scripts should be executed in the following order:

• intro.Rmd: An introduction, duh.
• preprocess.Rmd: Downloading the data and quality control
• cycle.Rmd: Cell cycle phase assignment
• normalize.Rmd: Calculation of cell-specific size factors
• variance.Rmd: Identification of highly variable genes
• dimred.Rmd: Dimensionality reduction with randomized PCA

Various output objects will be saved to objects/. A few of these objects are currently hosted at https://drive.google.com/open?id=1_0WbmJ2BriLKlyKEf1Bbb8K0_NwD9rw-. Note that sce.rds does not contain the actual counts or normalized expression values, and requires something like this:

library(TENxBrainData)
tenx <- TENxBrainData()
sce <- readRDS("sce.rds")
tenx <- tenx[,colnames(sce)] # drop 19,672 cells from the raw TENxBrainData
counts(sce) <- counts(tenx) # overwrite inbuilt absolute path
library(scater)
sce <- normalize(sce) # generate normalized expression values

The pics/make_pics.R scripts will generate the figures used in the paper.