Compute size factors based on the coverage of spike-in transcripts.
A SingleCellExperiment object with rows corresponding spike-in transcripts.
A character vector specifying which spike-in sets to use.
A string indicating which assay contains the counts.
A logical scalar indicating whether only size factors should be returned.
A logical scalar indicating whether the size factors should be stored for general use by all genes.
The size factor for each cell is defined as the sum of all spike-in counts in each cell. This is equivalent to normalizing to equalize spike-in coverage between cells. Size factors are scaled so that the mean of all size factors is unity, for standardization purposes if one were to compare different sets of size factors.
Spike-in counts are assumed to be stored in the rows specified by
This specification should have been performed by supplying the names of the spike-in sets – see
?isSpike for more details.
By default, if multiple spike-in sets are available, all of them will be used to compute the size factors.
The function can be restricted to a subset of the spike-ins by specifying the names of the desired spike-in sets in
By default, the function will store several copies of the same size factors in the output object.
One copy will be stored in
sizeFactors(x) for normalization of all genes – this can be disabled by setting
One copy will also be stored in
sizeFactors(x, type=s), where
s is the name of a spike-in set in
type=NULL, a copy is stored for every spike-in set, as all of them would be used to compute the size factors.)
Separate storage allows spike-in-specific normalization in
sf.out=TRUE, a numeric vector of size factors is returned directly.
Otherwise, an object of class
x is returned, containing size factors for all cells.
A copy of the vector is stored for each spike-in set that was used to compute the size factors.
general.use=TRUE, a copy is also stored for use by non-spike-in genes.
Lun ATL, McCarthy DJ and Marioni JC (2016). A step-by-step workflow for low-level analysis of single-cell RNA-seq data with Bioconductor. F1000Res. 5:2122
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
################ # Mocking up some data. set.seed(100) ncells <- 200 nspikes <- 100 spike.means <- 2^runif(nspikes, 3, 8) spike.disp <- 100/spike.means + 0.5 spike.data <- matrix(rnbinom(nspikes*ncells, mu=spike.means, size=1/spike.disp), ncol=ncells) ngenes <- 2000 cell.means <- 2^runif(ngenes, 2, 10) cell.disp <- 100/cell.means + 0.5 cell.data <- matrix(rnbinom(ngenes*ncells, mu=cell.means, size=1/cell.disp), ncol=ncells) combined <- rbind(cell.data, spike.data) colnames(combined) <- seq_len(ncells) rownames(combined) <- seq_len(nrow(combined)) y <- SingleCellExperiment(list(counts=combined)) isSpike(y, "Spike") <- ngenes + seq_len(nspikes) ################ # Computing and storing spike-in size factors. y2 <- computeSpikeFactors(y) head(sizeFactors(y2)) head(sizeFactors(y2, type="Spike")) # general.use=FALSE does not modify general size factors sizeFactors(y2) <- 1 sizeFactors(y2, type="Spike") <- 1 y2 <- computeSpikeFactors(y2, general.use=FALSE) head(sizeFactors(y2)) head(sizeFactors(y2, type="Spike"))
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.