VISION provides some convenience methods for loading gene expression data output from the 10x CellRanger pipeline.
The read_10x() and read_10x_h5() functions
load count data from 10x and perform the ID conversion from Ensembl IDs
to Gene Symbols. This is performed as most signature libraries refer to
genes using symbols. The expression of genes whose symbols map to
multiple Ensembl IDs are combined by summation. The resulting count
matrix can then be scaled and input into VISION:
# Read in expression counts (Genes X Cells)
counts <- read_10x(
expression = "outs/filtered_gene_bc_matrices/mm10/matrix.mtx",
genes = "outs/filtered_gene_bc_matrices/mm10/genes.tsv",
barcodes = "outs/filtered_gene_bc_matrices/mm10/barcodes.tsv"
)
# Scale counts within a sample
n.umi <- colSums(counts)
scaled_counts <- t(t(counts) / n.umi) * median(n.umi)
vis <- Vision(scaled_counts, signatures = c("data/h.all.v5.2.symbols.gmt"))